research on customer support

• SUGGESTED TOPICS
• The Magazine
• Newsletters
• Managing Yourself
• Managing Teams
• Work-life Balance
• The Big Idea
• Data & Visuals
• Reading Lists
• Case Selections
• HBR Learning
• Topic Feeds
• Account Settings
• Email Preferences

Customer service

• Sales and marketing
• Business management
• Customer experience

How Sales Reps Can Succeed in the Social Era

• April 10, 2013

Where Does the Customer Fit in a Service Operation?

• Richard B. Chase
• From the November 1978 Issue

The CEO of Ozon on Building an e-Commerce Giant in a Cash-Only Economy

• Maelle Gavet
• From the July–August 2014 Issue

Pepsi, United, and the Speed of Corporate Shame

• Andrew Winston
• April 12, 2017

Strong Patient-Provider Relationships Drive Healthier Outcomes

• Erin E. Sullivan
• Andy Ellner MD
• October 09, 2015

Customers' Revenge (HBR Case Study)

• December 01, 2007

Leading by Letting Go

• December 25, 2013

Competing on the Eight Dimensions of Quality

• David A. Garvin
• From the November 1987 Issue

The Future of Shopping

• Darrell K. Rigby
• From the December 2011 Issue

Listen to Your Frontline Employees

• Anthony K. Tjan
• April 04, 2012

Loyalty-Based Management

• Fred Reichheld
• From the March–April 1993 Issue

How Customers Can Rally Your Troops

• From the June 2011 Issue

Revolutionizing Customer Service

• Harvard Business Review
• From the April 2016 Issue

Retail Isn’t Broken. Stores Are

Customers' Revenge (Commentary for HBR Case Study)

• Nate Bennett
• Chris Martin
• Barak Libai

My Week as a Room-Service Waiter at the Ritz

• From the June 2002 Issue

Your Company Should Be Helping Customers on Social

• Maher Masri
• Dianne Esber
• Hugo Sarrazin
• Marc Singer
• July 15, 2015

Zero Defections: Quality Comes to Services (HBR OnPoint Enhanced Edition)

• W. Earl Sasser Jr.
• October 01, 2000

Trust Your Employees, Not Your Rule Book

• Bill Taylor
• April 20, 2017

Fighting Bias on the Front Lines

• From the November–December 2021 Issue

Customer-Introduced Variability in Service Operations

• Frances X. Frei
• March 14, 2006

Improving Customer Engagement

• Marcelo Olivares
• Andrew Shin
• October 19, 2011

HubSpot and Motion AI: Chatbot-Enabled CRM

• Thomas Steenburgh
• February 12, 2018

Pharmacy Service Improvement at CVS (A)

• Andrew McAfee
• December 14, 2005

Corey Robinson at Sprint Corporation (A)

• Linda A. Hill
• Lisa A. Pode
• Anthony J. Mayo
• June 08, 2005

The Mitchell Family and Mitchells/Richards

• Amy C. Edmondson
• Corey Hajim
• Kelly Mulderry
• John A. Davis
• November 08, 2004

Zappos.com 2009: Clothing, Customer Service, and Company Culture

• Robin J. Ely
• Laura Winig
• October 20, 2009

Rapid Rewards at Southwest Airlines

• August 11, 2001

Singapore Airlines: Customer Service Innovation

• Rohit Deshpande
• July 22, 2003

Buckman Laboratories (A)

• William E. Fulmer
• October 01, 1999

Chantale and Clinton Call for Service

• Christopher A. Ross
• March 25, 2009

Ritz-Carlton: Using Information Systems to Better Serve the Customer

• Thomas O. Jones
• Norman Klein
• October 20, 1994

Designing Transformational Customer Experiences

• Stefan Thomke
• March 06, 2017

Beleza Natural

• Omar Besbes
• Nelson M. Fraiman
• Maria J. Quinteros
• Gabriel Y. Weintraub
• July 16, 2012

Willow Creek Community Church (B): Cracks In the Foundation

• Leonard A. Schlesinger
• January 21, 1999

Corey Robinson at Sprint Corporation (B)

Cvs, jon roberts and josh flum, video.

• March 01, 2007

Redefining the AXA Brand

• Stefan Michel
• Jean-Pierre Baillot
• April 12, 2010

Managing a Global Team: Greg James at Sun Microsystems, Inc. (A)

• Tsedal Neeley
• Thomas J. DeLong
• July 03, 2008

Building Brand Community on the Harley-Davidson Posse Ride (Multimedia Case)

• Susan Fournier
• Sylvia Sensiper
• James McAlexander
• John Schouten
• December 28, 2000

Making Customer-Centric Strategies Take Hold

• June 02, 2015

v12 Delivery Services - LT 500 users

• July 01, 2014

Delivering Excellent Service: Lessons from the Best Firms

• Robert Ford
• Cherrill P. Heaton
• Stephen W. Brown
• October 01, 2001

Popular Topics

Partner center.

Everything You Need to Know About Customer Experience Research

Updated: January 20, 2023

Published: October 27, 2022

Think back to the last time you received amazing customer service . Remember how it made you feel and how you perceived that business before and after your experience. Compare that experience to the last negative encounter you had with a business, and the difference could not be more obvious.

With recent CX trends such as omni-channel marketing and support, along with the continued growth of e-commerce, it's necessary for companies to understand the customer experience (CX) from multiple angles to reduce pain points and improve customer satisfaction.

CX is not something that your company can just ignore, as nearly half of all customers report that CX is more important to them in 2021 than it was just a year ago. Given this surge in demand for a quality experience, how can your company pivot to meet your customers' rising expectations?

The answer lies in conducting extensive customer experience research. Keep reading to learn everything you need to know about CX research, or use the links below to jump ahead:

What is customer experience research?

Why is customer experience research important, customer experience research tips, customer experience research methods, start conducting your own customer experience research.

Customer experience is the summation of every interaction that a customer has with your company throughout their journey. From a cold call to a service inquiry or a coupon in the mail, each interaction between your company and a customer helps to create individual impressions, perceptions, and behaviors that together make up the customer experience.

Meanwhile, customer experience research represents the actionable steps that your company can take to understand CX. This includes collecting customer data — both pre-and post-sale — and then analyzing that data for trends that can lead to process, product, or service improvements.

Best practices in customer experience research programs include focusing on three core components:

• Development

Image source

Your company's CS research journey starts with a customer experience strategy that lays out your vision of your company's goals and maps out the customer journey as it stands and how you hope it to be.

Once you have a strategy in place, you can then put your ideas into action and develop tools and practices for measuring, organizing, and deciphering the data you'll need to validate any changes you make.

Finally, the research process ends with the tracking and implementation of findings that your company can use as a foundation for continuous improvements to CX design.

Customer Satisfaction vs. Customer Experience

To truly understand CX research, we must first take a moment to differentiate customer experience from customer satisfaction. Although the two terms are often used interchangeably, they are actually quite different and should not be conflated with one another.

Customer satisfaction is a measurement used to gauge how happy a customer is with your company's products, services, or brand overall.

It pays to have happy customers, with 89% of consumers admitting that they are more likely to make an additional purchase after a positive customer service experience.

While customer satisfaction aims to measure how a customer feels about your company — whether good, bad, or neutral — customer experience attempts to measure every interaction that your customers have throughout their entire relationship with your company.

Customer experience research can help you tease out key CX data points and measure your company's success against them. A few of those data points are highlighted in the image below.

All of these metrics and more combine to make up the customer experience. With carefully planned and executed customer experience research, your company can glean insights from these interactions that you can then use to enhance your CX design and raise client satisfaction.

There's nothing worse than losing a customer to a competitor due to a poor experience. Unfortunately, this reality is all too common, with 58% of American consumers reporting that they will switch companies because of a negative customer service experience.

Regardless of the industry, CX is highly correlated with brand loyalty, with the customers reporting the most positive experiences also scoring highest on surveys measuring brand loyalty.

On average, there is a 38% difference in likelihood to recommend a company between customers that rated a company's CX as "good" versus customers that rated that company's CX as "poor."

The ROI of conducting customer experience research is well worth the expense, especially when you stop to consider the alternatives.

After all, it's well known that lead generation is one of the most daunting tasks faced by any company. Yet, at the same time, it costs between 5 and 25 times more to acquire a new customer than to retain an existing one.

It's no wonder that 48% of customer service professionals state that creating a positive customer experience is a top priority for their team.

There are as many methods to conduct customer experience research as there are ways that customers interact with businesses.

Some companies will choose to use deductive reasoning and use commonly held assumptions and perceptions from the market and their customers to map out the customer experience and make changes from there.

On the other hand, other companies will opt to use inductive reasoning and take small sample sets of observable data and use that information to create their CX map and inform their decision-making.

Whatever route your company chooses, it's important to drill down and identify the essential aspects of what you're hoping to gain from this research.

The questions highlighted in the image below are a great place to start.

These questions and more need to be addressed before your company attempts to analyze a shred of evidence. If you skip the planning and strategizing phase of the CX research process, then you're doomed to fail before you begin, because your company won't know what customer experience research questions it's trying to answer.

Once you've settled on your questions, it's time to start organizing the tools and resources you'll need to actually conduct your research.

Customer Experience Research Tools and Resources

Depending on your goals, you may choose to collect qualitative data that provides in-depth CX insights. However, this type of data is not easy to quantify. For example, long-form customer interviews provide a wealth of information about how customers see your CX but the results are difficult to reduce to actionable insights.

Alternatively, your company may decide to focus on measuring and tracking CX key performance indicators and highlight the collection of quantitative data. Surveys are one of the most commonly used mediums to collect quantitative data, as they allow companies to easily sort and organize responses into groups that can be used for statistical analysis and comparison.

Whatever customer experience research method your company chooses, it's essential that leadership is all on the same page to embrace CX research as a key aspect of your business. With as many as 93% of CX initiatives destined to fail, you want to make sure you're doing everything you can to make sure the time you're investing into CX research is well-spent and not just more money down the drain.

Traditionally speaking, most customer experience research was carried out by large marketing research firms that conducted the interviews, focus groups, and surveys that companies used to make changes to their CX design.

Today, the research landscape also includes data collection firms that help companies collate and store their data for easy retrieval and analysis.

That said, many companies also choose to conduct their own research in-house using a variety of research methods for collecting, organizing, and interpreting data.

As shown in the image above, some of the most common methods of collecting CX research data include:

• Feedback Software

Let's discuss each in more detail.

1. Interviews

Interviews provide a wealth of qualitative data, while surveys are highly customizable, allowing your company to tailor its surveys to collect any type of quantitative data. However, these methods are often more time-consuming and labor-intensive than other methods, so are usually conducted by larger organizations with more resources and time.

Two of the most popular surveys are also among the easiest methods of conducting CX research: NPS and CSAT.

Net promoter score (NPS) is a benchmark used to determine how likely a customer is to recommend your business to someone. NPS surveys are useful, as they measure how a customer feels overall about your brand, which allows your company to gather lots of big-picture information.

Then there's customer satisfaction score (CSAT), which measures customer satisfaction with a particular interaction, product, or service. CSAT surveys allow your company to get quantifiable data concerning every little detail of your business that can then be used to design specific solutions.

3. Feedback Software

In addition, many companies now turn to feedback software to help them collect, organize, and track CX data from multiple sources. These applications make it easy for companies to conduct CX research by bringing sophisticated analysis software and technology support all within one system.

Each type of CX method provides valuable information to the table that your company can use to improve the customer experience. Still, you'll need to make sure that you're following CX research best practices to ensure that you get the most out of your efforts.

Customers are no longer willing to settle for a bad shopping experience to get the best price or a superior product.

The new normal requires successful companies to be sensitive to their customers' needs and smooth pain points when and where they emerge. To do this, companies need to invest in CX research that paints a portrait of the customer journey, identifies areas of improvement, and urges leadership to implement actionable changes.

If your company is serious about prioritizing the customer experience, then you need to do the requisite research. That way, you can turn your assumptions into meaningful solutions that let your customers know you care about them.

And we all know there's nothing better than a satisfied customer.

Don't forget to share this post!

Related articles.

How AI Image Misuse Made a World of Miscommunication [Willy's Chocolate Experience]

7 Ways to Delight Your Customers This Holiday Season

14 Customer Experience Fails that Companies Can Learn From

How Customer Experience Has Evolved Over the Last Decade [+ 2024 Trends]

Memorable Examples of AR in Customer Experience [+Tips for Implementing the Technology]

How to Create an Effective Customer Journey Map [Examples + Template]

Digital Customer Experience: The Ultimate Guide for 2023

How to Implement a Hybrid Customer Service Strategy That Works [Expert Tips]

User Flows: 8 Tips For Creating A Super Smooth User Experience

11 Best Practices for B2B Customer Experience

Outline your company's customer journey and experience with these 7 free customer journey map templates.

Service Hub provides everything you need to delight and retain customers while supporting the success of your whole front office

• Browse All Articles
• Newsletter Sign-Up

CustomerSatisfaction →

No results found in working knowledge.

• Were any results found in one of the other content buckets on the left?
• Try removing some search filters.
• Use different search filters.

The next frontier of customer engagement: AI-enabled customer service

How to engage customers —and keep them engaged—is a focal question for organizations across the business-to-consumer (B2C) landscape, where disintermediation by digital platforms continues to erode traditional business models. Engaged customers are more loyal, have more touchpoints with their chosen brands, and deliver greater value over their lifetime.

About the authors

This article is a collaborative effort by Avinash Chandra Das, Greg Phalin , Ishwar Lal Patidar, Malcolm Gomes, Rakshit Sawhney, and Renny Thomas , representing views from McKinsey’s Operations Practice.

Yet financial institutions have often struggled to secure the deep consumer engagement typical in other mobile app–intermediated services. The average visit to a bank app lasts only half as long as a visit to an online shopping app, and only one-quarter as long as a visit to a gaming app. Hence, customer service offers one of the few opportunities available to transform financial-services interactions into memorable and long-lasting engagements.

Those customers are getting harder to please. Two-thirds of millennials expect real-time customer service, for example, and three-quarters of all customers expect consistent cross-channel service experience. And with cost pressures rising at least as quickly as service expectations, the obvious response—adding more well-trained employees to deliver great customer service—isn’t a viable option.

Companies are therefore turning to AI to deliver the proactive, personalized service customers want, when and how they want it—sometimes even before they know they want it. For transformed organizations, AI-enabled customer service can increase customer engagement, resulting in increased cross-sell and upsell opportunities while reducing cost-to-serve. In global banking alone, research from McKinsey conducted in 2020 estimates that AI technologies could potentially deliver up to $1 trillion of additional value each year , of which revamped customer service accounts for a significant portion. 1 “ AI bank of the future: Can banks meet the AI challenge ,” McKinsey, September 19, 2020.

While a few leading institutions are now transforming their customer service through apps, and new interfaces like social and easy payment systems, many across the industry are still playing catch-up. Institutions are finding that making the most of AI tools to transform customer service is not simply a case of deploying the latest technology. Customer service leaders face challenges ranging from selecting the most important use cases for AI to integrating technology with legacy systems and finding the right talent and organizational governance structures.

But done well, an AI-enabled customer service transformation can unlock significant value for the business—creating a virtuous circle of better service, higher satisfaction, and increasing customer engagement.

Would you like to learn more about our Operations Practice ?

The perils and promise of ai customer engagement.

Multiple converging factors have made the case for AI-based customer service transformation stronger than ever. Among the most important: increased customer acceptance of (and even preference for) machine-led conversational AI interactions. Meanwhile, related technologies such as messaging platforms are becoming more accessible, and customer behaviors are becoming more understandable with the relentless expansion of data pools institutions can collect and analyze.

Three challenges

But challenges also loom. First, complexity . The COVID-19 pandemic acted as a major catalyst for migration to self-service digital channels, and customers continue to show a preference for digital servicing channels as the “first point of contact.” As a result, customers increasingly turn to contact centers and assisted-chat functions for more complicated needs. That raises the second issue: higher expectations . Customer confidence in self-service channels for transactional activities is leading them to expect similar outcomes for more involved requests. Businesses are therefore rapidly adopting conversational AI, proactive nudges, and predictive engines to transform every point of the customer service experience. Yet these moves raise demand for highly sought-after skills, generating the third challenge: squeezed labor markets that leave customer service leaders struggling to fill crucial roles.

How leaders fulfill AI’s customer engagement promise

What ai-driven customer service maturity looks like.

A few leading institutions have reached level four on a five-level scale describing the maturity of a company’s AI-driven customer service.

Level 1: Manual and high-touch, based on paper forms and offered largely via assisted channels.

• Reactive service, with the majority of interactions on human-assisted channels
• Paper use is still prevalent

Level 2: Partly automated and basic digital channels, with digitization and automation of servicing in assisted channels.

• Reactive service, with limited self-servicing opportunities
• Lower adoption of available self-service channels
• Lower availability of digital or straight-through-processing (STP)

Level 3: Accessible and speedy service via digital channels, with self-servicing on select channels and a focus on enabling end-to-end resolution.

• Somewhat proactive, but limited engagement
• Self-service channels such as mobile apps, interactive voice response (IVR) systems, and internet sites handle half of all interactions, and can support STP.

Level 4: Proactive and efficient engagement deploying AI-enabled tech, with self-servicing enabled by proactive customer interactions and conversational user experience (UX).

• Proactive, with high customer engagement on digital channels
• Self-service channels such as mobile apps, IVR systems, and internet sites handle 70-80 percent of interactions and can support most requests and transactions

Level 5: Personalized, digitally enabled engagement, bringing back the human touch via predictive intent recognition.

• Engagement via service interactions that are personalized and proactive at the individual customer level
• Digital touchpoints drive service-based engagement, for example via enhanced cross-selling and upselling
• More than 95 percent of service interactions and requests can be solved via digital and STP channels

Leaders in AI-enabled customer engagement have committed to an ongoing journey of investment, learning, and improvement, through five levels of maturity. At level one, servicing is predominantly manual, paper-based, and high-touch. At level five—the most advanced end of the maturity scale—companies are delivering proactive, service-led engagement, which lets them handle more than 95 percent of their service interactions via AI and digital channels (see sidebar, “What AI-driven customer service maturity looks like”).

The most mature companies tend to operate in digital-native sectors like ecommerce, taxi aggregation, and over-the-top (OTT) media services. In more traditional B2C sectors, such as banking, telecommunications, and insurance, some organizations have reached levels three and four of the maturity scale, with the most advanced players beginning to push towards level five. These businesses are using AI and technology to support proactive and personalized customer engagement through self-serve tools, revamped apps, new interfaces, dynamic interactive voice response (IVR), and chat.

Myth busters: Unexpected insights on contact centers

Toward engaging, ai-powered customer service.

To achieve the promise of AI-enabled customer service, companies can match the reimagined vision for engagement across all customer touchpoints to the appropriate AI-powered tools, core technology, and data. Exhibit 1 captures the new model for customer service—from communicating with customers before they even reach out with a specific need, through to providing AI-supported solutions and evaluating performance after the fact.

The human factor in AI-supported service

AI-powered does not mean automation-only. It’s true that chatbots and similar technology can deliver proactive customer outreach, reducing human-assisted volumes and costs while simplifying the client experience. Nevertheless, an estimated 75 percent of customers use multiple channels in their ongoing experience . 2 “ The state of customer care in 2022 ,” McKinsey, July 8, 2022. A reimagined AI-supported customer service model therefore encompasses all touchpoints—not only digital self-service channels but also agent-supported options in branches or on social-media platforms, where AI can assist employees in real time to deliver high-quality outcomes.

Even before customers get in touch, an AI-supported system can anticipate their likely needs and generate prompts for the agent. For example, the system might flag that the customer’s credit-card bill is higher than usual, while also highlighting minimum-balance requirements and suggesting payment-plan options to offer. If the customer calls, the agent can not only address an immediate question, but also offer support that deepens the relationship and potentially avoids an additional call from the customer later on.

AI service in the field: an Asian bank’s experience

Put together, next-generation customer service aligns AI, technology, and data to reimagine customer service (Exhibit 2). That was the approach a fast-growing bank in Asia took when it found itself facing increasing complaints, slow resolution times, rising cost-to-serve, and low uptake of self-service channels.

Over a 12-month period, the bank reimagined engagement. It revamped existing channels, improving straight-through processing in self-service options while launching new, dedicated video and social-media channels. To drive a personalized experience, servicing channels are supported by AI-powered decision making, including speech and sentiment analytics to enable automated intent recognition and resolution. Enhanced measurement practices provide real-time tracking of performance against customer engagement aspirations, targets, and service level agreements, while new governance models and processes deal with issues such as service request backlogs.

Underpinning the vision is an API-driven tech stack, which in the future may also include edge technologies like next-best-action solutions and behavioral analytics. And finally, the entire transformation is implemented and sustained via an integrated operating model, bringing together service, business, and product leaders, together with a capability-building academy.

The transformation resulted in a doubling to tripling of self-service channel use, a 40 to 50 percent reduction in service interactions, and a more than 20 percent reduction in cost-to-serve. Incidence ratios on assisted channels fell by 20-30 percent, improving both the customer and employee experience.

Seizing the opportunity

To leapfrog competitors in using customer service to foster engagement, financial institutions can start by focusing on a few imperatives.

• Envision the future of service, keeping customers and their engagement at the core while also defining the strategic value to be attained—for example, a larger share of wallet with existing customers? Expansion of particular services, lines of business, or demographics?
• Rethink every customer touchpoint, whether digital or assisted, together with opportunities to enhance the experience while also increasing efficiencies.
• Maximize every customer service interaction, to deepen customer relationships, build loyalty, and drive greater value over the customer’s lifetime.
• Leverage AI and an end-to-end technology stack, to provide a more proactive and personalized customer service experience that supports self-service and decision-making for customers as well as employees.
• Adapt agile and collaborative approaches to drive transformation, comprised of SMEs from different business and support functions of the organization.

Holistically transforming customer service into engagement through re-imagined, AI-led capabilities can improve customer experience, reduce costs, and increase sales, helping businesses maximize value over the customer lifetime. For institutions, the time to act is now.

Avinash Chandra Das is an associate partner in McKinsey’s Bengaluru office, where Malcolm Gomes is a partner and Ishwar Lal Patidar is an expert. Greg Phalin is a senior partner in the Charlotte office, Rakshit Sawhney is an associate partner in the Gurugram office, and Renny Thomas is a senior partner in the Mumbai office.

The authors wish to thank Amit Gupta, John Larson, and Thomas Wind for their contributions to this article.

Explore a career with us

Related articles.

Smart scheduling: How to solve workforce-planning challenges with AI

17 Ways to Conduct Customer Research Right Now & Collect Valuable Feedback

Table of contents

Whether you’re marketing a brand new startup or a seasoned veteran, there’s no substitute for real customer feedback and research.

After all, you can’t market anything effectively if you don’t know who you’re selling to.

Customer research is such a crucial part of marketing that, when we asked survey respondents how important they considered customer research to be, nearly 93% rated it as “Very Important” or “Crucially Important.”

“Marketers need to conduct customer research at the very least annually. In order to sell to someone, you need to know their needs,” said Tim Brown of Hook Agency .

Brown’s comment got us thinking—if customer research is so important, how often are people doing it? When we asked those same marketers that question, we got some varied responses. But crucially, the majority skewed toward more often, with over 25% reporting quarterly customer research efforts and nearly 20% reporting they conduct customer research daily .

So what are marketers actually doing to conduct that customer research? When we asked our respondents about that, there were 4 clear winners that more than half of the marketers we spoke with reported using:

• Customer interviews
• Email surveys
• Analytics analysis
• Online research

But we also heard about many other creative ways to conduct customer research that we hadn’t thought of before.

On that note, here are the 20 customer research methods marketers shared with us.

1. Leverage Existing Customer Reviews

Brian Jensen of Congruent Digital recommended turning to a familiar source for customer research: online reviews. “We used a tool called Apify to crawl and return all of our client’s reviews into a database. We then put into a text analysis tool to find the top keywords and phrases (attributes) customers used in their reviews.”

Jensen says they used this data to help improve the client’s messaging.

“Once we had the data and knew by occurrences what their customers enjoyed most about their experience, we updated ads and landing pages to better identify with the needs and expectations of prospects.”

2. Spend a Day in Your Customer’s Office

Phil Strazzulla of SelectSoftware shared another customer research method we hadn’t heard about before. Strazzulla recommended spending a full day, in-office with your customer, saying “This allows me to have informal conversations with the key stakeholders I need to market to in order to better understand their challenges, goals, language, and personalities.”

“Simply reach out to a potential or current customer and ask if you can work from their space for a day,” Strazzulla explained. “And have as much free time as you can to walk around and talk to people in the office about what they do and how you can help them with your product.”

3. Turn to Data Analytics

Analytics analysis was one of the top 4 answers we heard—but it’s a broad term, so we were interested to learn more about what marketers do with analytics.

“When we do customer and product research, we start by understanding how customers are using the tool by looking at their data and usage, and then benchmarking it with their industry,” said Supratim Da Dam of CallPage . “This allows us to have a solid idea of how our customers are deploying our solution, the gaps, successes, blockers, and more.”

Robert Baillieul of Lombardi Publishing uses Twitter Analytics to identify topics and pains that resonate with their customers. “Anything that consistently generates engagement rates north of 5% indicates a huge pain point for your customer—sometimes issues they would never admit to out loud. You can then turn these insights into new products, services, or content.”

“We get data from many tools we’re using (email marketing, website analytics, social media, and more),” explained Jonathan Aufray of Growth Hackers . “With the help of a great data analyst and a tool like Google Data Studio, we can quickly analyze our customers.”

Vira Vielmann of Seventh Scout says they turn to social media analytics most often. “We typically utilize social media analytics to learn more about the audience engaging with us. This gives us an amazing insight into their demographics and interests. They also let us know what topics and posts are doing well and which aren’t performing the best, so we can adjust our strategy and editorial calendars as needed.”  

4. Collect Customer Survey Responses

“My favorite way to get customer research is to send out an email survey,” James Pollard of The Advisor Coach said. “I keep it short (about five or six questions) and only ask them the questions that will have the biggest impact on my business.”

Based on the marketers we spoke with, there are more benefits to this type of research to learn the voice of customer than you may expect.

“When you really pay attention to the way that people share information with you,” Amber Vilhauer of NGNG Enterprises said, “you’ll notice your audience using specific verbiage and wording that you can bake into your website. Often times the way that you would describe your services is very different than the way that a customer or prospect would describe those services.”

“Ultimately, people want good products that will serve them well,” Mr. SR of Semi-Retire Plan explained, “so they do have an interest in giving you helpful information to improve your —especially if they’re an existing customer who already has an affinity for your brand.”

That said, marketing consultant Farheen Gill suggests giving customers a little added incentive. “Include them in the last phase of your welcome email journeys, but also offer giveaways for other surveys you need to run throughout the year (i.e. ‘Respond today to be entered into a drawing for a $50 gift card’).”

“What’s important,” said Andrea Loubier of Mailbird , “is that you dig deep with your surveys. Asking generic questions isn’t going to get you very far. Make sure your multiple-choice questions offer diverse answers and don’t be afraid to ask the hard questions. You may be shocked at just how much your customers are willing to share.”

Louis Watton of Shiply suggested another tip for getting insightful, honest answers. “One creative approach we’ve used in customer research is not letting interviewees know the company conducting the research at first.”

Explaining, Watton added, “Often we’ve found that customers will hold back on criticism if they know you work for that company. The most valuable insights and potential improvements we’ve learned have come from asking broader questions about the industry, which allows them to talk freely without worrying about insulting anyone.”

“We launch every new survey or questionnaire with a video,” said Charles Musselwhite of FunLovingCouples . “We don’t ask any more than 12 questions at a time, and we always add in a weird and obscure question or two to keep people on their toes and engaged.”

5. Watch Customers Use Your Product

Samuel Wheeler of Inseev Interactive offered up another top-notch tip, recommending marketers actually watch customers using the product, navigating the website, interacting with content, and more.

“It’s a great idea to ask users to narrate their thought process as they navigate the page and ask them to actually take an action (purchase or form submission). In addition to asking the users to talk through their decision-making process.”

“It’s a great way to get both quantitative and qualitative data,“ Wheeler added.

If you need to understand how customers are using your product to gather feedback, one tool you should consider for customer feedback is Usersnap. This helpful tool allows product managers, software engineers, designers, and marketers to instantly collect information from users on-site through screen captures, screen recordings, surveys, feature requests, menu buttons, in-app forms, visual drawings, and bug reports.

Another  feedback tool  you might consider to crowdsource customer feedback and feature requests is UseResponse. This tool allows you to create feedback communities where customers can post their feedback, while others can comment and upvote it.

Pro Tip: Here Is Your Go-To Dashboard For Measuring the Performance of Your Customer Support Team

No matter your role in customer support – agent, manager, or VP – your core focus is to ensure that customers’ issues, complaints, and information requests are always dealt with promptly and efficiently. But to stay on track, you may have to spend hours manually compiling data from different tools into a comprehensive report. Now you can quickly monitor and analyze your customer service performance data from Intercom in a single dashboard that monitors fundamental metrics, such as:

• New conversations . Track the total number of new conversations your customer support team handles daily, weekly, monthly, or within the specified date range.
• Open conversations by team member . View the total number of conversations in your support inbox that are still open and find out which team members are handling them.
• Leads . Track the number of leads generated by your customer support team within a specified date range. Dig deeper to learn the nature of the messages that help convert visitors to leads, and use your insights to improve future conversations.
• Users by tag name . View the total number of conversations your customer support team has handled over time and see how your team members tagged those messages in Intercom. Using tags makes it easier for anyone monitoring the dashboard to learn more about customer needs, interests, and issues.

Now you can benefit from the experience of our customer support experts, who have put together a plug-and-play Databox template that contains all the essential metrics for monitoring and analyzing the performance of your customer support reps. It’s simple to implement and start using as a standalone dashboard or in customer service reports, and best of all, it’s free!

You can easily set it up in just a few clicks – no coding required.

To set up the dashboard, follow these 3 simple steps:

Step 1: Get the template 

Step 2: Connect your Intercom account with Databox. 

Step 3: Watch your dashboard populate in seconds.

6. Leverage Publicly Available Data

We talk a lot about gathering and analyzing data these days, but one thing marketers often forget about is the wealth of existing data that are publicly available online. “A lot of people overlook the incredible amount of data that the government and nonprofits collect that can be useful for customer research,” said Jeromy Sonne of Reverb Agency .

“The most creative approach I’ve used to learn more about my customers is public records, which give me additional information about the customer’s location, demographics, behavioral specialties,” added Emily Andrews of RecordsFinder . “Public records have a big database, which helps me to understand how better I can sell my clients’ goods or services.”

Carmine Mastropierro of Mastro Commerce told us about a hybrid customer research process: “One approach I’ve used to learn more about my customers is a mix of online research and market research tools.”

“Studying industry reports,” Mastropierro explained, “allowed me to get a broad overview of who my customers are and how they behave. Then, Google Analytics and other online tools helped me narrow down demographics, interests, and other behaviors to refine my audience.

7. Use Facebook Audience Insights

Casey Hill of Bonjoro also recommended pulling customer data from where it’s readily available already. In Hill’s case, that’s Facebook’s Audience Insights tool.

“It’s a free tool through Facebook,” Hill explained, “and it will give you information on any intended audience.” According to Hill, Audience Insights can help marketers answer questions like:

• What kind of jobs do customers have?
• When are they active online?
• What pages do they follow?

“It’s an incredible tool for customer research that many people aren’t aware exists.”

8. Have a One-on-One Conversation

“I find that doing a 30-minute video call beats every other type of research,” said Corey Haines of Hey Marketers . “With the right questions in hand and a friendly conversational tone, so much can be uncovered that you would never know otherwise.”

Sarah McIntyre of Bright Inbound Marketing agreed with Haines, saying, “Actually talking with people is critically important to understand, not just what they think about your product or service, but how they found you, what the sales process was like, who else they were considering, why they chose you. Unless you actually ask, you’ll be running your marketing based on assumptions.”

According to Renee Bauer, Hello Marketing Agency abides by a similar strategy for customer research. “We do regular NPS surveys for a client, and we ask responders to let us know if they are willing to participate in a one-on-one interview. These interviews serve as a helpful supplement to persona research, and provide actionable information for our client about what’s important to their current customers and how they need to improve their service.”

“Face to face encounters in a more social setting (as opposed to an interview or focus group) will give you the most honest, instinctive, and digestible feedback,” said Kyle Turk of Keynote Search.

“Online feedback methods, although they still provide great feedback, allows the user to spend too much time thinking of a response, and the ability to manipulate their responses. It also really only captures your promoters and detractors. The core customer group that is neutral about your product or service will not engage in the feedback, leaving a large gap in data.”

Anna Kaine of ESM Inbound echoed Turk, noting that “picking up the phone for a talk with customers is always more personal and genuine than just sending out a questionnaire—because you can really probe and show you’re listening. It’s a far more human experience.”

“We are clear and open about the focus of the calls, and they’re always happy to help us – after all, it’s in their best interests for us to focus closer on their pain points,” Kaine added.

Paige Arnof-Fenn of Mavens & Moguls recommended make a tour of customer interviews. “Go on a Listening Tour. Ask a few smart, open-ended questions, then sit back and take notice. Start listening with no strings attached and you’ll be amazed at what you find.”

Ever Increasing Circles ’ Alistair Dodds seconded Arnof-Fenn’s last point, adding, “We’ve found out things that I don’t think would have ever come up in an office or business environment. And it’s helped us to really focus in on how to get the client to their real objective.”

9. Conduct Research With Google

It’s no surprise that the king, queen, and jester of online research is, of course, Google. But the marketers we spoke with noted so many novel ways to use Google search for customer research, including:

• Reading competitors’ customer reviews on Google My Business
• Researching the way customers speak about your product and industry
• Tailoring content toward real customer pain points and questions

“Google is an excellent resource to learn more about your customers, without the use of expensive tools,” said Ben Johnston of Sagefrog Marketing Group . “If you’re in a competitive space, look at your competition’s Google My Business profiles and read the reviews of satisfied and unsatisfied customers to learn what real customers like or don’t like about your direct competition.

Roman Zhyvitsk of Travel SEO Agency touted the importance of using Google to better understand how your customers speak about (and search for) your business. “When you sell your products or services online, it is highly important to know what search phrases people use to find it. Very often it is not as obvious as you might think.”

Johnston also noted how Google can help with ensuring content resonates with your customers, saying, “You can refine your content ideas to actually engage with your customer base by looking at ‘People Also Ask’ or ‘People Also Search For.’ That’s a direct insight into what kinds of questions your customer base is asking and what they’re interested in.”

Set Up Google Alerts for Customers and Prospects

In addition to conducting manual customer research on Google, Carlos Puig of BUNCH shared another pro tip: Google Alerts.

“Right after signing a contract with a new customer, I strongly recommend setting up a Google Alert for the name of the company and the names of the people you closed the deal with. Google will keep pushing relevant information that will help you understand the situation of your client and detect potential upsells.”

10. Ask Customers to Rate Your UX

Much of the advice we heard focused on overall customer information. But Victor Antiu of Sleek Bill says they focus on the micro aspects of customer experiences, too.

“Throughout the app, we marked micro-conversions. When the user finishes one (for example creates and sends an invoice), we show a small rating bar and based on the score he gives us, we either show him a small survey to find out what was hard, or we thank him and ask what we can improve.”

“It’s a similar system to what Skype and Booking.com do,” Antiu explained. “It’s a simple way to find pain points or issues in various funnels.”

11. Use Social Listening

“Social media is probably the best tool that you could use to understand the thought process of your client,” said Harry Gandia of Igniting Movement . “Social media can help a marketer discover what their target audience is thinking in real-time. Not many other mediums can offer that. And it’s totally free.”

Many of the marketers we spoke with invoked one form of social listening or another. After all, social media is where customers hang out—regardless of who your customers are.

Find Their Online Groups and Hangouts

“One approach we use to learn more about our customers,” Kelsey Miller of Pepperland Marketing explained, “is to find the online groups, forums, and communities that they frequent. This can be in the form of Facebook groups, Reddit threads, industry-specific forums, hashtags, and so on. This is helpful in understanding how these people interact with each other, the questions they are asking, the challenges they are facing, and so much more.”

Alexandra Sheehan of Coach Content recommends turning to Facebook Groups specifically. “I love joining Facebook groups that my audience is likely to be a part of and just observing their behavior. This shows you what really makes them tick. The things that annoy them, their true pain points, their sense of humor, little nuances like that.”

“ Find out where your customers are hanging out online,” advised Vinoth AJ of Apoyo Corp , adding, “One proven method is Quora. All we have to do is type a topic and it will display all questions related to that topic. Go ahead and read all the questions related to your market.”

Create Your Own Group

While many marketers recommend going where the customers are, there’s also some benefit to taking the Field of Dreams approach.

“By far the best way to learn more about our customers has been to create a dedicated Facebook group around our products,” said Jonathan Chan of Insane Growth . “Not only does this give us the ability to foster a real sense of community around our brand, but we have routine access to the most highly-engaged members of our audience.”

Jack Paxtone of VYPER echoed Chan, explaining, “Hosting a forum either on our website or on Reddit turned out to be a great way to build a database of feedback from our clients, while also engaging with them to build a strong relationship for our brand.”

“The Vyper Facebook Group is currently our most popular platform for getting to know our customers,” Paxtone added. “We can freely interact with each other, understand their likes and dislikes, and also request valuable feedback when we are beta testing new products and services.”

Jarrod Miller-Dean of Housecall Pro added, “We utilize community outreach in our private Facebook group. For example, by posing a question in the group and asking members for their help and response.”

12. Use Heatmap Tools to Understand How Customers See Your Website

Customer research is about more than just who your customers are. It’s also about understanding how they interact with your brand and your product. That’s why Sneh Ratna Choudhary of Beaconstac recommended using a heat mapping tool to better understand and optimize their website for the customers visiting.

“We’ve been using Hotjar to understand the exact pain points of users to implement a human-centered design.”

“For instance,” Choudhary explained, “our free QR Code Generator tool was receiving visitors, but there weren’t any real conversions. We looked at Hotjar videos only to find out that we had way too many CTAs to begin with. Upon realizing this, we scaled down our CTA to include only 3 major CTAs and our visitor-to-trial conversion rate is currently hovering at 15.6%.”

13. Keep It Informal

For some, customer research can feel like a weighty, formal undertaking—but it doesn’t have to be, and many of the marketers we heard from reminded us of that.

“So many business owners and entrepreneurs think that market research is this big, complicated thing,” Carla Williams Johnson of Carli Communications pointed out. “And, while you can conduct structured surveys and questionnaires, you can also simply ask your customers directly what they think of an idea that you may have.”

“Sometimes that direct, informal approach can give you the best feedback,” Johnson added.

Liz Courtney of BBMG took that idea to the next level, saying, “To get more realistic and meaningful insight into consumers’ needs, aspirations and behavior, we try to connect with them on their own turf. Visiting them in their homes, going shopping with them, or chatting with them in pairs with a friend rather than forcing them into unnatural settings like sterile focus groups or relying only on multiple-choice surveys.”

14. Tap Your Network for Feedback

Kathleen Marrero of First Fig Marketing & Consulting emphasized the effect an existing relationship can have on the kind of customer research and feedback you end up with, suggesting your network is a great place to start.

“I have found the best way to learn more about potential customers is to open up a friendly dialogue with connections I have on social media platforms. I have reached out to numerous connections on sites like LinkedIn and asked for a real, honest conversation about whatever space I am gathering information within, the good and bad and any other information that would help me better serve the community.”

“I have found that people are very willing to offer insight if there is no sales pitch,” Marrero added.

15. Leverage Your Email Subscribers

“Reaching out to email subscribers to ask what’s bothering them is one of the most effective ways to learn more about customers,” said Priscilla Tan of Content Kapow .

“Two weeks ago,” Tan shared, “I was struggling to write a blog post. I didn’t know which topics to focus on. Rather than going with my gut, I asked my subscribers. I gave them 3 options and picked the one with the most number of votes. Not only did it help with topic development, but it also helped me to dig deeper into the pain points they’re facing at work.”

16. Offer a Beta Version of Your Product in Exchange for Feedback

One common thread throughout the responses we heard was that, while customers do have an incentive to help you create a better product for them, that isn’t always enough to entice feedback or survey responses.

To combat that problem, Carsten Schaefer of Crowdy.ai suggested offering a beta or paired-down version of your product in exchange.

“We launched a beta for 100 days before going live with our product. We gave our beta users all the features completely free in exchange for one thing: feedback about our product and how they used it for their business,” Schaefer explained. “It has brought us incredible insights which we used in the final iteration of the product.”

17. Learn from Live Chat and Support Interactions

If there’s one painfully overlooked source of customer research, it’s the support team. Few other teams within a business have the kind of direct contact with customers that customer support pros see every day.

Zack Naylor of Aurelius said, “I make it a point to answer every single live chat we get on our website for product questions and requests. Often what happens is that I get to learn a lot about potential customers from what they’re looking for and end up being able to schedule a live call to dive deeper and learn more.”

Get to Know Your Customers

Customers are the lifeblood of every successful business, and finding business traction and growth depends on your ability to get to know and understand your customers.

Whether you’re ready to go big with a large, organized customer survey or simply want to chat one-on-one with a few customers, you’ll emerge better equipped to serve their needs and grow the business.

Get practical strategies that drive consistent growth

Google My Business SEO: 8 Tips to Optimize Your Business for Local Searches

The 23 Best Keyword Tracking Tools (According to 107 SEOs)

23 Ways to Improve Your Landing Page Conversion Rates

Build your first dashboard in 5 minutes or less

Latest from our blog

• Playmaker Spotlight: Tory Ferrall, Director of Revenue Operations March 27, 2024
• New in Databox: Safeguard Your Data With Advanced Security Settings March 18, 2024
• Metrics & KPIs
• vs. Tableau
• vs. Looker Studio
• vs. Klipfolio
• vs. Power BI
• vs. Whatagraph
• vs. AgencyAnalytics
• Product & Engineering
• Inside Databox
• Terms of Service
• Privacy Policy
• Talent Resources
• We're Hiring!
• Help Center
• API Documentation

Just one more step to your free trial.

.surveysparrow.com

Already using SurveySparrow?  Login

By clicking on "Get Started", I agree to the Privacy Policy and Terms of Service .

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

Don't miss the future of CX at RefineCX USA!  Register Now

Enterprise Survey Software

Enterprise Survey Software to thrive in your business ecosystem

NPS Software

Turn customers into promoters

Offline Survey

Real-time data collection, on the move. Go internet-independent.

360 Assessment

Conduct omnidirectional employee assessments. Increase productivity, grow together.

Reputation Management

Turn your existing customers into raving promoters by monitoring online reviews.

Ticket Management

Build loyalty and advocacy by delivering personalized support experiences that matter.

Chatbot for Website

Collect feedback smartly from your website visitors with the engaging Chatbot for website.

Swift, easy, secure. Scalable for your organization.

Executive Dashboard

Customer journey map, craft beautiful surveys, share surveys, gain rich insights, recurring surveys, white label surveys, embedded surveys, conversational forms, mobile-first surveys, audience management, smart surveys, video surveys, secure surveys, api, webhooks, integrations, survey themes, accept payments, custom workflows, all features, customer experience, employee experience, product experience, marketing experience, sales experience, hospitality & travel, market research, saas startup programs, wall of love, success stories, sparrowcast, nps benchmarks, learning centre, apps & integrations, testimonials.

Our surveys come with superpowers ⚡

Blog Best Of

Customer Research 101: Definition, Types, and Methods

12 February 2024

Table Of Contents

What is Customer Research?

Why is customer research important, types of customer research.

• 6 Customer Research Methods
• How SurveySparrow Can Help

Do you want to improve your marketing or product? Then, customer research can help.

Your customer is at the heart of all your business decisions. In fact, everything revolves around a customer. A business is about having a paying customer, and it wouldn’t exist without one.

The effectiveness of your product or marketing depends on how well you know your customers. When you know your customers better, you can make better product or marketing decisions.

In this article, we break down:

• What customer research is
• Why it’s valuable for your business
• Different types of customer research
• Six customer research methods you can use to refine and grow your business

Customer research (or consumer research ) is a set of techniques used to identify the needs, preferences, behaviors, and motivations of your current or potential customers.

Simply put, the consumer research process is a way for businesses to collect information and learn from their customers so they can serve them better.

Businesses typically conduct customer research to uncover new insights on their customers. They then use these newly uncovered insights to improve their product, craft an effective marketing strategy, and more.

Here are 2 key questions customer research helps you answer:

• Who are my ideal customers? Who is the best fit (or worst fit) for our product?
• What channels can I use to find and communicate with my ideal customers?

Online survey tools like SurveySparrow can help you answer these questions. With omnichannel survey distribution, snazzy data visualization, and 1,500+ integrations with your favorite tools, SurveySparrow simplifies customer research for your GTM and product teams.

Please enter a valid Email ID.

• 14-Day Free Trial
• • Cancel Anytime
• • No Credit Card Required
• • Need a Demo?

A. How well do you know your customers? Not knowing enough about your customers can cost you time and money.

For example, a recent survey revealed that 46% of customers broke up with a brand because they received irrelevant content pushes.

Successful marketers realize that research is necessary to understand and cater to the ever-changing needs of today’s customers. According to a study by Coschedule:

• Successful marketers are 242% more likely to conduct audience research at least once every quarter.
• 56% of the study’s most elite marketers research at least once a month.

B. You shouldn’t make assumptions about your customers’ preferences or needs. You have to go out there and get opinions from real customers.

C. You need to go beyond your general idea about your customers. The more you understand your customers, the better you’ll be able to serve them with your product or service.

D. If you want to make your product the best in the market, you need to identify any unmet needs and learn how well your product serves the needs of your current customers.

E. Customer research helps you learn more about your customers, both the potential and existing ones. Serving your customers better than the alternatives starts with understanding them better and more deeply.

F. Here are other key reasons why you should research customers:

• Know the Why : Your analytics dashboard merely tells you what your customers do. Only research can help you understand why they do that.
• Validate Assumptions and Best Practices : In most cases, guesswork leads to terrible decisions. Your customers might not need what you think they need. And what works for most businesses might not work for you. The only real way to know is to talk to your customers.

Customer research can be done in two distinct ways: primary and secondary.

Primary research

Primary research is research you conduct yourself. In other words, in primary research, you collect the data yourself. Some examples of primary research are face-to-face interviews, surveys, and social media interactions.

Secondary research

Secondary research (or desk research ) is done by someone else. In secondary research, you make use of data that’s been collected by other people. A few examples of secondary research are forums or communities, industry reports, and online databases.

Primary and secondary research can be further broken down into two kinds of data: qualitative and quantitative.

Qualitative data

Qualitative data is descriptive and conceptual. And the nature of the data makes it subjective and interpretive. Examples of qualitative data include descriptions of certain attributes, such as blue eyes or chocolate-flavored ice cream .

Quantitative data

Quantitative data can be expressed using numbers, which means it can be counted or measured. As opposed to qualitative data, it’s objective and conclusive. Examples of quantitative data include numerical values such as measurements , length , cost , or weight .

Customer Research Methods that Work in 2024 (and Beyond)

Now that you know what customer research is and why it’s important, read on to learn the different consumer research methods you can use to make the most of it.

In a survey, you ask a series of questions to your customers regarding a subject or concept.

You can conduct a survey in person, over the phone, through emails, or online forms.

Here are some advantages of conducting customer research through surveys:

• Quickly collect a ton of insightful data without the high costs.
• The data you collect using surveys is simple to analyze.
• You can ask various questions since you get a wide range of question formats.

When it comes to surveys, it’s all about how you ask. Clear and concise questions can help you get reliable information.

An online survey tool is your best bet for quickly gathering customer information. All you need to do is create a survey with a ready-to-use template and send your customers a link to take it.

If you’re in need of a cost-free and easy-to-use solution for conducting customer research surveys and beyond, consider exploring SurveySparrow . This tool aids in gathering essential data by enabling you to conduct thorough data analysis via its user-friendly and conversational survey format.

Check Out SurveySparrow for Free here! 

14-Day Free Trial • No Credit Card Required • No Strings Attached

In an interview, you speak directly to your customers and ask them open-ended questions.

• Interviews allow you to have deep, one-on-one conversations with your customers and explore a topic in-depth.
• You can go into the details, obtain data beyond surface-level information, and gather deeper insights.

While interviews allow you to probe deeper into a subject, success depends on the expertise and skills of the researcher (or interviewer) conducting the interviews.

Conducting interviews isn’t easy. It’s time-consuming and costly. However, the information you collect can be invaluable for your company’s growth.

You can meet your customers in person to conduct your interviews. Or you can use video conferencing tools such as Google Meet or Zoom to converse with your customers online.

Your analytics dashboard lets you in on your customers’ actions within your product.

Just a glance at it and you’ll know what your customers do and how they engage with your product.

The irony is that customers don’t know what they want or why. They might think they need something but that might not be the case.

What they say they need doesn’t equate to what they do.

The point is that customer-reported behavior is different from actual behavior. That’s why it pays to track and observe your customers’ behavior.

You can use heatmaps, click tracking, scroll mapping, and user-recorded sessions to gain insights into your users’ actions and behavior.

Focus Groups

In this method, you combine a small group based on certain criteria such as demographic, firmographic, or behavioral attributes.

And you ask this group about whatever topic or concept. It could be about your product, marketing message, or something else that’s related to your customers or business.

The idea is to get them to talk to each other and have meaningful conversations.

A moderator helps facilitate the conversations between the individuals in this group. The moderator will try to draw meaningful insights from these conversations and discussions.

You mainly use this technique to understand a certain topic or subject better.

Competitive Analysis

Studying your competitors’ strategies and tactics is a great way to learn more about the target market and the existing solutions.

You can analyze both your direct and indirect competitors depending on the needs you address and the customers you cater to.

You can conduct a competitive analysis from a marketing or product perspective.

If you conduct your analysis from a marketing perspective, you study your competition’s SEO strategy , landing page copy, blog content, PR coverage, social media presence, etc.

You can also conduct your competitive analysis from a product perspective and analyze your competitors’ user experience, features, pricing structure, etc.

Review Mining

The reviews of you and your competitors are another great way to get inside your customer’s head. This method can be especially valuable if you are a SAAS company.

It helps you better understand your competitor’s strengths and weaknesses as well as your own. This understanding helps you improve your own products and better address the needs of your ideal customers.

This kind of data is easy to acquire as it’s publicly available, and you can get them on:

• Review sites such as G2Crowd and Capterra.
• Forums and niche communities such as ProductHunt, Reddit, Quora, etc.

Why SurveySparrow is the Best Customer Research Tool

SurveySparrow facilitates comprehensive customer research by enabling businesses to efficiently collect, analyze, and act on customer feedback, leading to better informed and customer-centric decisions.

• Collect Feedback Easily : Create simple surveys to find out what customers think about your products or services.
• Understand Satisfaction : Use surveys to figure out how happy customers are with what you offer.
• Learn Buying Habits : Find out why customers buy certain products, which helps in planning what to sell.
• Get Product Opinions : Ask customers what they like or don’t like about your products to make improvements.
• See How People View Your Brand : Understand how customers see your brand, which is important for your marketing.
• Keep Up with Trends : Regular surveys help you stay updated on what your customers want or need.
• Group Customers : Identify different types of customers to target them more effectively with your marketing.
• Improve Customer Experience : Learn where you can make the buying process better for your customers.
• Test New Ideas : Before launching new products, check if your customers would be interested.
• Check Customer Loyalty : Find out if customers would keep using your products or recommend them to others.

Sign up for a free trial.

  please enter a valid email id. signup for free 14-day free trial • no credit card required • no strings attached, final thoughts.

Businesses that deeply understand their customers have a huge advantage over the ones that don’t. Period.

Whatever you’re looking to learn or achieve, it becomes a lot clearer with a little research.

When done right, customer research can be your competitive advantage.

Be sure to pick a method that’s right for your situation. What are you looking to learn and achieve? Think through each research method carefully and pick the one that works best for you.

Have you conducted customer research? What did you learn? And how did it go? Tell us about that in the comment section below.

And if you’re looking to conduct customer research through surveys, feel free to check out SurveySparrow .

I'm a developer turned marketer, working as a Product Marketer at SurveySparrow — A survey tool that lets anyone create beautiful, conversational surveys people love to answer.

You Might Also Like

The 2020 beginner’s guide for net promoter score calculation, how to become a ceo: a complete guide, 20+ best nps survey questions and templates (2024 guide), cherry-picked blog posts. the best of the best..

Leave us your email, we wont spam. Promise!

Start your free trial today

No Credit Card Required. 14-Day Free Trial

Request a Demo

Want to learn more about SurveySparrow? We'll be in touch soon!

Research Your Customers with Conversational Surveys!

Create beautiful surveys that your customers will love to answer boost your response rates by 40%.

14-Day Free Trial • No Credit card required • 40% more completion rate

Hi there, we use cookies to offer you a better browsing experience and to analyze site traffic. By continuing to use our website, you consent to the use of these cookies. Learn More

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Behav Sci (Basel)

Providing Excellent Customer Service Is Therapeutic: Insights from an Implicit Association Neuromarketing Study

Gemma anne calvert.

1 Nanyang Business School, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Singapore; gs.ude.utn@CAMIL

Abhishek Pathak

2 School of Business, 4 Nethergate, University of Dundee, Dundee DD1 4HN, UK; [email protected]

Lim Elison Ai Ching

Geraldine trufil.

3 Split Second Research Limited, London E1 8FA, UK; [email protected] (G.T.); [email protected] (E.P.F.)

Eamon Philip Fulcher

This paper reports the results of a combined biometric and implicit affective priming study of the emotional consequences of being the provider or receiver of either positive or negative customer service experiences. The study was conducted in two stages. Study 1 captured the moment-by-moment implicit emotional and physiological responses associated with receiving and providing good customer service. Study 2 employed an affective priming task to evaluate the implicit associations with good and poor customer service in a large sample of 1200 respondents across three Western countries. Our results show that both giving and receiving good customer service was perceived as pleasurable (Study 1) and at the same time, was implicitly associated with positive feelings (Study 2). The authors discuss the implications of the research for service providers in terms of the impact of these interactions on employee wellbeing, staff retention rates and customer satisfaction.

1. Introduction

Customer satisfaction is a vital goal for all businesses because it leads to increased sales and customer re-patronage, which ultimately boosts profits. To this end, managing customer experiences across the customer–employee touchpoints plays a critical role, given that most businesses involve some level of direct contact (e.g., face-to-face or voice-to-voice) between employees (especially those working at the consumer interface) and customers. Yet delivering high quality and effective customer service is not a straightforward or easily managed process. Customer–employee interactions have a significant emotional component that often confounds training strategies. While it is understood that positive customer service results in better marketing outcomes, much less is known about the emotional impact on those responsible for delivering that service.

Service employees often hide their true inner feelings and maintain a pleasant facial and bodily display in a bid to please their customers and/or gain control over employee–customer interactions [ 1 , 2 ]. Indeed, companies often train their service employees to act in a friendly manner since the display of positive emotions is associated with favourable consequences, such as increased customer satisfaction, customer re-patronage, and positive word-of-mouth [ 3 ]. Such acting requires significant effort on the part of the employees and can cause employees to suffer emotional burn-out if they are required to “put on” displayed emotions for long periods of time [ 4 , 5 ]. Furthermore, consumers do not always appreciate employee friendliness, and may even construe it negatively as being disrespectful [ 6 ]. Indeed, consumers are increasingly adept at discerning the expressive behaviour of service providers. For instance, they are more likely to be moved by the authenticity of an employee’s smile rather than the extent of it [ 7 , 8 ]. The somewhat artificial nature of these exchanges, coupled with the constant requirement to suppress negative emotions and “appear” friendly and understanding, makes it extremely difficult to disentangle true emotions associated with positive and negative customer–staff interactions and those which individuals presume they should experience.

Measuring the emotional consequences associated with customer experience is further complicated by the fact that it involves multiple moments of contact between an organisation and a customer. These may include the feelings evoked when walking into a shop, the way in which the customer is treated by frontline service employees in-store, as well as post-purchase follow-up customer service. Furthermore, the ability of individuals to introspect and comment on the nature of these subjective emotional responses, particularly during dynamic social interactions, is highly variable and often inaccurate [ 9 , 10 ]. The relationship (and perceptions thereof) between the employee-customer interaction has traditionally been measured using surveys [ 11 ] (may not be accurate always, and we propose an alternate method in this the current paper.

Extant scholarly researchers as well as companies interested in assessing service quality mostly employ explicit, self-reported measures. However, this approach captures only a partial picture of the multitude of responses in consumers’ brains. Neuroscience research has shown that a vast amount of human behaviour is driven and influenced by emotional and cognitive responses that occur below conscious awareness [ 12 , 13 , 14 ]. At the conscious level, customers tend to know what they want and also how they wish to be treated. But important implications of good and poor customer service can also play out at the subconscious or implicit level of cognition [ 15 , 16 ]. The same is true for those responsible for providing that service, where multiple conscious and subconscious emotional factors impinge on the effectiveness of customer interactions.

Although it is well known that the quality of customer–employee interaction is crucial for organisations and the importance of customer service has been studied for many years now, the literature is scarce on the consequences of poor (or good) service on employees [ 17 , 18 ]. Poor customer–employee interaction can lead to employee stress and is a potential health risk [ 19 ], which can cost up to $ 300 billion in losses cumulatively to organisations the world over American Institute of Stress (2014). Employees who are regularly tasked to maintain positive interactions with customers have also been reported to show excessive emotional burden, exhaustion and absenteeism [ 20 , 21 ]. Similarly, customer mistreatment (and consequent stress) can compromise both short term and long-term employee well-being [ 22 ] and result in emotional exhaustion [ 23 ].

Recent research has also shown that positive customer behaviour during service interactions has a cross over positive effect on the employee [ 24 ]. Similarly, stressful customer interactions can have a negative impact on the affective state of employees [ 25 ].

In order to develop a deeper understanding of the implicit consequences of customer service on providers and receivers, this research examined the implicit emotional responses associated with receiving and providing excellent service. Specifically, the paper investigated 1) the perception of both giving and receiving good vs. bad customer service, 2) and the implicit associations (or feelings) which people associate with the experience of giving or receiving good vs. bad customer service. By doing so, this research contributes to the services literature by demonstrating how the positive benefits of excellent customer service can impact not only on customers, but also on service providers themselves. Such positive outcomes, if made explicit, can clearly be exploited in a positive way so as to increase employee job satisfaction and reduce staff turnover rates. Furthermore, this research also contributes to the field by proposing a new research approach that captures customers’ subconscious responses in order to gain a more comprehensive understanding of the subliminal effects of positive customer–employee interactions.

2. Background

Over the past decade, techniques that have emerged from the fields of neuroscience and psychology, such as functional MRI, electroencephalography (EEG), eye-tracking, biometrics, facial decoding and implicit association testing, have been engaged by brand owners to capture these vital subconscious responses in order to define and predict consumer behaviour with much greater accuracy (for a recent review, see [ 9 ]). This approach has been referred to as “neuromarketing” [ 26 ] and numerous commercial practitioners of this burgeoning industry now exist. In recent years, commercial entities have paid particular attention to neuromarketing methods that are scalable, cost-effective and offer fast turnaround times [ 27 ].

One methodology that satisfies these criteria is the use of implicit reaction time tests [ 28 ]. The mainstay of many cognitive psychology experiments since the 1970s, implicit reaction time paradigms measure individuals spontaneous or ‘gut instinct’ responses. Commercial adaptations of these paradigms permit marketers to capture these vital subconscious consumer responses online, without the need for verbal feedback or even respondents’ awareness of their reactions. Implicit measures have now been used in a variety of settings to extract people’s implicit emotions and attitudes to a wide range of different issues, including racial prejudice, sexual preferences, alcoholism, mental health, and consumer attitudes (see [ 29 ] for an overview). Importantly, the implicit responses obtained in these studies were shown to be more predictive of respondents’ subsequent behaviour than their explicit verbal responses obtained at the same time and are therefore, in many instances, more accurate indicators of their emotional responses to specific concepts and scenarios.

Several recent implicit reaction time paradigms have been shown to have high reliability and validity [ 30 , 31 , 32 , 33 ]. These approaches rely on a simple behavioural response—a very rapid key press to the presentation of a stimulus, which is made following a simple decision about the stimulus. There are several distinct implicit paradigms, each with specific strengths and weaknesses, and the choice of task depends on the research question being addressed [ 29 ].

In the current study, we employed two implicit reaction time tests. The first was the Impulse Test recently developed and shown to measure the emotions evoked as respondents view dynamic material (e.g., while watching a television advertisement, movie trailer or video footage [ 34 ]. The second was an affective semantic priming task [ 35 , 36 ] that assesses the strength of implicit association between a set of emotional words and specific concepts, in this case, good and poor customer service. The rationale for employing two distinct implicit tests was that in the first case, we were able to identify the immediate emotions elicited by positive customer service interactions (both from the perspective of the provider and the receiver) and relevant in short-term memory, and in the second case, we were able to capture the more deep-seated emotions associated with positive as well as negative customer interactions that are stored in long-term memory.

Physiological responses (heart and breathing rate and electrodermal changes) were also measured during the Impulse test to determine if positive customer service interactions (both providing and receiving) impact the levels of arousal. Arousal, one of the components of emotional responding, is associated with stress, anxiety and fear [ 37 ], and physiological manifestations of arousal include increased blood pressure, heart rate, sweating and hyperventilation [ 38 ]. We hypothesized that the act of simply observing positive customer–staff interactions would result in reduced arousal and therefore stress levels, similarly to that experienced when engaging in other everyday pleasures.

This study was conducted in two stages. Study 1 was designed with two objectives in mind: first, the study served to identify the nature of the immediate emotions elicited in real time as respondents viewed videos of people receiving or providing excellent customer service compared with viewing other positive scenarios (e.g., everyday pleasurable activities such as enjoying time with friends), and secondly, we wanted to examine the physiological responses (heart and breathing rate, and electrodermal response) to the customer service scenarios depicted in the videos. In Study 2, we examined the more deeply held emotions (i.e., those maintained in long term memory) associated with customer service interactions (positive and negative) in a larger population (N = 1200) across three countries that individuals have either delivered or received.

3.1. Study 1: Laboratory Based Study

3.1.1. participants.

Twenty participants (thirteen females (two left-handed) and seven males (all right-handed) with mean age of 27 years) were recruited from Bristol, UK (via flyers in exchange for vouchers) and given small incentives to take part in a study to measure their immediate physiological and psychological responses to different emotional scenarios in real-time, including footage depicting individuals providing or receiving customer service (sample size is similar to other studies of comparable nature, e.g., [ 39 ]).

3.1.2. Materials

Three distinct video clips, each one minute in duration, were professionally created specifically for this study:

Video 1 (Condition 1: Control): was made up of footage of everyday pleasures (unrelated to customer service) shown from the first person perspective, such as eating crisps, going for a walk in the park.

Video 2 (Condition 2: Providing excellent customer service): constituted footage of four different scenarios in which service staff were filmed delivering excellent customer service and the footage shown from the service provider’s perspective. The scenarios were as follows: (i) a booking agent is seen giving a customer tickets to a previously sold out play at the theatre and knowing she has had a hard time recently, the booking agent has gone one step further and arranged for her to go to the opening night party as well, (ii) a travel agent helps a couple, who have been separated for six months due to work, to plan their dream honeymoon, giving them personalised recommendations on where to visit and restaurants to eat out at, (iii) a groom leaves his wedding rings in the back of a taxi the day before the wedding. The taxi driver returns to the hotel where he dropped off the groom off, re-uniting him with the rings and thus saving the day, and (iv) a woman collapses in a restaurant while on holiday after which a fellow customer, a doctor, tries to help but her friend is very distressed and does not speak the local language. The waitress steps in to translate what the doctor is saying and accompanies them all to hospital.

Video 3 : (Condition 3: Receiving excellent customer service): shows scenarios featuring excellent customer service and are the same scenarios as those used in Condition 2 but re-filmed and shown from the customer’s perspective.

3.1.3. Protocol

Only one subject at a time participated in the experiment. Each participant was greeted by the experimenter who explained that the purpose of the study was to gain a better understanding of customer service interactions. After obtaining informed consent (FREC-EF02-PSY-16-1-2013), physiological electrodes were applied and subjects were seated in front of a computer screen. Heart and breathing rate as well as skin conductance measures were collected as subjects viewed the video clips. The experimental videos were shown on a computer screen and participants’ responses were recorded using the computer keyboard. The order of presentation of the three videos was counterbalanced across subjects.

BIOPAC physiological equipment was used in the collection of data. In order to measure heart rate, one electrode was placed on the medial surface of each leg just above the ankle. A third electrode was placed on the right anterior forearm at the wrist. Once electrodes were attached, participants were asked to remain still while the system parameters were calibrated. Data was recorded at a rate of 200 times a second. Heart rate was measured as the milliseconds between heart beats and was analyzed as the average heart-rate per two seconds (400 datapoints).

Skin conductance data was collected through two electrodes attached to the middle and ring finger of the non-dominant hand. The index finger was avoided as it was used in the reaction time task. Participants were given the opportunity to practice key pressing with minimal movement of the hand, so as not to disturb recording. Respiratory cycle was recorded through a respiratory transducer attached around the chest below the armpits and over the shirt. It was adjusted so that it was slightly tight at the point of maximal expiration.

During the acquisition of biometric data, subjects were also asked to carry out an implicit reaction time test (the Impulse test) while viewing the experimental videos. The Impulse test consists of two stages—a baseline phase and an experimental phase. In the baseline phase of the current task, participants were exposed to a set of emotional words (see Table 1 , presented one at a time and in randomised order in the centre of the computer screen). Each word was presented four times, and on each occasion, the words were displayed on the screen until the correct key was pressed or 2 s had elapsed. The next word was presented 2 s after the previous word. The selection of emotional words most relevant to the content shown in our three videos was determined in a prior pilot study in which 16 words (8 positive and 8 negative) were identified from a cohort of 50 words as being ranked most closely to the emotions elicited in the videos and categorised consistently as of positive or negative valence.

Emotional words used during the Impulse test.

On each trial, an emotional word appeared briefly on the computer screen and subjects were instructed to categorise them according to their emotional valence by pressing the “I” key on the keyboard if the word was positive in nature and the “E” key if the word was negative (key mapping was counterbalanced between the subjects). A reminder of which key corresponded to which emotional valence “Positive” or “Negative” remained on the computer screen in the top left and right corners throughout the two phases. Subjects were asked to respond as quickly and as accurately as possible.

The baseline trials and the experimental trials were identical, except that a video was played in the background during the experimental trials. The baseline phase of the Impulse test served both as a means of training the subjects to respond within a very short timeframe (to clear contamination from conscious brain processes) and also to familiarise the participants with the task. Responses that were deemed too slow to be classified as pre-cognitive were followed by a brief warning tone and the visual warning “too slow”. Following successful completion of the training phase, subjects were informed via instructions on the computer that the experimental phase was about to begin. The design was similar to the practice phase, however, during this phase, the words appeared superimposed on the dynamic footage (the three videos are described in the Materials section). Respondents were instructed to continue classifying the words as positive or negative in terms of emotional valence and to press the corresponding keys on the computer keyboard.

Our previous research has shown that the speed and accuracy of classification of these words reflect the feelings that a participant has towards the content of the movie or television clip [ 34 ]. By comparing reaction times to classify positive and negative emotional words during the training (baseline) and experimental phase, it is possible to infer the nature of the internal feelings elicited in the viewer by the video content every 2 s. Specifically, we have found that positive emotions elicited by the footage shown, speeds responses to positive words and slows them to negative ones. The reverse holds true for aspects of the footage that elicits negative emotions—responses to negative words are sped up and responses to positive words are slowed. In both cases, RTs were computed against the responses recorded during the baseline and training phase, for each participant. To understand this approach in the context of semantic priming studies, in the current study, the video content acts as the “priming” stimulus, with the emotional words being the targets.

3.1.4. Analysis

The physiological analysis focused on the emotional peaks detected whilst watching each video. We hypothesized that during each emotionally provocative video, there were likely to be fluctuations in arousal, as determined by changes in electro-dermal response, breathing rate and heart rate. We also hypothesized that these physiological indices would be accompanied by changes in implicit psychological emotional responses detected using the Impulse test.

Physiological Responses

For the physiological data, the maximum values for each physiological measure (heart rate, breathing rate and skin conductance) were first computed by extracting the peak response recorded every 2 s during both the training and experimental phase of the Impulse test and the results from all individuals averaged and tested for statistical significance using paired T-tests at each time point (see Figure 1 ).

This graph shows the heart rate of one representative participant when they were carrying out the baseline test (blue line) and the test with the movie clip in the background (red line). The resulting data computed for this participant is the difference between the blue and red heart rate values every two seconds. When the value of a point on the red line is larger than the value of the corresponding point on the blue line (e.g., at t = 6), it shows that the participant’s heart rate increased as a result of watching this part of the movie clip. Conversely, at t = 28, the participant’s heart rate shows a decrease. These values were computed for each participant and then averaged and subjected to statistical analysis.

Impulse Test

Reaction time data obtained during the training and experimental phases of the Impulse test were first subjected to pre-processing, including removal of outliers so that responses that were impossibly quick (<250 ms) and those that were so slow as to invoke conscious processing (>1200 ms), were removed. The data were then analysed following the method outlined by Fazio and Olson [ 36 ]. For all trials of each word presented in each video, a facilitation index (FI) was computed. For all congruent responses (e.g., classifying “delightful” as “Positive”) obtained for each word across all trials, the FI was computed by subtracting the reaction times during the experimental phase from those obtained during the baseline phase. For incongruent responses to each word (e.g., classifying the word “lonely” as “Negative”), the FI calculation was reversed such that reaction times obtained during the baseline phase were subtracted from the reaction times obtained during the experimental phase. Thus, an FI greater than zero implies a response that is congruent with the emotion word set and a FI less than zero implies a response incongruent (or opposite) with the word set. This approach allowed us to take into account both the congruency (or subjective accuracy) of responses as well as the reaction times. The dependent variable was, for each moment of each video (every two seconds), the percentage of participants whose FI indicated that the footage at each time-point was either congruent or incongruent with a positive emotion or negative emotion. The averaged data were then tested for each condition for statistical significance using the binomial test that computes the probability of obtaining a specific count in one direction (e.g., a positive emotional response) against the total number of observations (the number of positive and negative responses).

3.1.5. Results

Physiological measures.

Condition 1: (Everyday pleasures) While viewing a video depicting everyday pleasures, breathing rate dropped from 16.3 cycles to 15.4 cycles per minute); heart rate remained stable at 76.1BPM in both cases; and a non-significant increase in electrodermal response from 0.171 to 0.252 was recorded.

Condition 2: (Providing good customer service) was associated with an average increase in heart rate from 76.0 BPM during the baseline phase to 87.4 BPM while the video was shown in the background ( p < 0.01). Breathing rate decreased from 16.7 cycles per minute during the baseline to 10.2 cycles per minute while viewing the video, and a significant increase in electrodermal response from 0.114 to 0.335 ( p < 0.001) was recorded.

Condition 3: (Receiving good customer service). A statistical comparison of physiological measures revealed that viewing footage of others receiving excellent customer service resulted in a significant increase in electrodermal response from 0.164 to 0.308 ( p < 0.01) and a significant decrease in heart rate from 71.4 to 80.6 BPM ( p < 0.05). There were no significant differences in breathing rate for condition 3 (17.2 to 16.8).

The control condition (viewing everyday pleasures) elicited an FI of −3.15, showing that that there was a slightly shorter mean response latency to negative attributes than to positive attributes. However, this FI did not differ from zero ( p > 0.05). Viewing footage of individuals receiving excellent customer service elicited an FI of +36.9, which reveals a significant increase in response latency to positive attributes ( p < 0.001). Viewing footage of individuals providing excellent customer service yielded the largest increase in FI of +53.8 ( p < 0.001). Paired t-tests revealed that providing excellent customer service elicited a greater association with positive emotions than either receiving excellent service ( p < 0.05) or viewing everyday pleasures ( p < 0.001).

3.1.6. Discussion and Conclusion

We believe that this is the first demonstration that viewing instances of positive customer service interactions from the perspective of both the recipient and the service provider has a positive impact on physiology and emotional well-being. Specifically, viewing footage of people delivering or receiving excellent customer service resulted in a significant increase in arousal levels, as evidenced by the increase in galvanic skin response and a significant decrease in heart rate (compared to viewing scenarios of everyday pleasures), indicating that positive customer service interactions can have a stress-reducing and calming impact on the service provider and surrounding viewers.

The results of the Impulse reaction time study showed that participants were faster at correctly classify positive word targets than negative ones when viewing footage of people providing good service compared to receiving it, or while viewing footage of every day pleasures. This is an intriguing finding as we would have hypothesized that people would adopt a self-interested stance and would instinctively attach greater positive valence to receiving good service than watching examples of people providing good service. Receiving good service was perceived with the same level of positive emotional engagement as viewing every day pleasures, highlighting the growing significance of customer service in people’s lives today.

The results of Study 1 raised further questions relating to the generalizability of these findings across different countries, age groups and gender. Therefore, in the next study, we sought to extend these findings by investigating the implicit emotional feelings associated with both positive, as well as negative, customer service interactions in a larger population using a web-based implicit affective priming task designed to uncover the strength of emotional association that people hold about positive and negative customer service interactions.

3.2. Study 2: Online Study

3.2.1. participants.

Participants (N = 1200) from three countries (UK, Canada and Australia; N = 400 from each country, 50% males) were recruited through a research participation recruitment company (Research Now) and were given small incentives to complete the tests. All the participants had normal to corrected vision, were native English speakers between 18 and 60 years and completed an online consent form prior to participation (the sample size is adequate for the chosen experimental design, since the study is a four (providing excellent or poor customer service vs. receiving excellent or poor customer service) by two (excellent service vs. poor service) design and is similar to other studies of comparable nature, e.g., [ 35 , 40 ].

3.2.2. Materials

The web-based survey included three components: (i) demographic questions to confirm age, gender, handedness and previous employment in a service industry, (ii) a consent form, and (iii) an implicit affective priming task. The survey was programmed in Javascript so that as soon as participants entered the survey, the test would automatically and immediately be downloaded onto their pc/laptop so that reaction times could be captured using the internal timing devices on the pc/laptop, which are far more sensitive than if running a program of this nature across the internet. On completion of the survey, the individual datasets were then uploaded back onto the server for analysis and without being apparent to the participant.

The affective priming task consisted of a series of emotional word primes ( Table 2 ) and target statements ( Table 3 ). The emotional word primes were selected following an explicit pilot test in 150 people (50 from each country) during which respondents were asked to classify attributes (from a set of 60; including those used in Study 1) into those most likely to be experienced in the context of extremely pleasurable experiences, peace of mind experiences, everyday experiences, and negative experiences. Of these, 35 were consistently classified and used as primes in the implicit test. The brief statements used as targets (e.g., “being helpful”, “feeling relieved”) were generated in consultation with service industry consultants and refer to the behaviours that were most often experienced in the context with excellent or poor service scenarios ( Table 3 ).

Emotional prime words used in the affective priming task.

Emotional words used to create brief statements used in Test A (Providing) and Test B (Receiving). All target words using in Test A were presented prefixed with the word “being” (e.g., “being helpful”, “being friendly”), whereas those used for Test B were pre-fixed with the word “feeling” (e.g., “feeling relieved”, “feeling neglected”).

3.2.3. Protocol

On entering the survey, respondents were asked to confirm their age, gender and handedness. They were also asked if there were currently employed, and/or did voluntary work and whether their current or any past employment involved “providing service of some form to service users, such as clients, customers or patients”. If they answered “no” to the last question, they were thanked for participating but informed that they were not eligible for the study.

On completing the inclusion criteria questions and subsequent consent form, participants were then asked to classify each of the 35 emotion words (pre-selected for inclusion in the implicit test) as extremely pleasurable experiences, peace of mind experiences, mundane experiences and negative experiences.

Participants were then instructed that they would be asked to perform a reaction time task that would measure how quickly and accurately they could classify a series of short phrases (see Table 3 ) that would be presented in the centre of the computer screen. There were two tasks designed to identify emotions implicitly associated with providing excellent or poor customer service (Test A) or receiving excellent or poor customer service (Test B). Participants were randomly assigned to one of the two tasks.

Before the experimental trials, participants were given 24 practice trials during which they were asked to discriminate whether short phrases which were either positive or negative (e.g., “being helpful”, “being impolite” in the case of test A— providing excellent or poor service) and (e.g., “feeling special”, “feeling neglected” in the case of test B— receiving excellent or poor customer service) were synonymous with either “excellent service” or “poor service” and to press the “E” or “I” key on the computer keyboard corresponding to each option. The practice trials served as a learning phase during which respondents were able to learn the association between each target and the correct key press so that they would not need to focus on which key to press during the main test.

The keys were allocated to “excellent service” or “poor service” and were counterbalanced for each participant, and once assigned, remained so for the duration of the task. If a response was incorrect, the error message “Try again!” appeared near the lower part of the screen; if two keys were pushed at the same time, the message “Please press only one key at a time” was displayed; if no key was pushed within two seconds, the cue “Warning: Please press E or I” appeared. The next trial proceeded after a 1500 ms inter-trial interval. Participants were instructed to respond as quickly as possible but to avoid making a mistake.

Following practice trials, participants were told that the main trials were about to begin and would be very similar as the practice phase but this time a word or “prime” was presented for 500 ms, immediately before the short phrase targets. Each prime was presented four times in total, twice before a phrase associated with “excellent service” and twice before a phrase associated with “poor service” to ensure a sufficient number of trials of each type. Prior testing has shown that with an N = 1200, this number of trials is sufficient to be able to detect a statistical difference if it exists. The task was identical to that conducted during the practice trials, which was to discriminate whether the targets (e.g., “being” or “feeling” a positive or negative emotion) that appeared immediately after the primes (see Table 2 ) were associated with “excellent service” or “poor service” and to respond as quickly and as accurately as possible by pressing the key corresponding to each option.

3.2.4. Analysis

Data were first subjected to analysis to remove outliers, including response times that were impossibly fast (<250ms) or those that occurred after the permitted time window. Reaction times were then computed for each word attribute and for each participant. A difference score was computed being the mean reaction time when the prime was presented before ‘poor service’ minus the mean reaction time when the prime was presented before ‘excellent service’. This was also done separately for Tests A (Providing) and B (Receiving). Positive difference scores indicated that a prime was more strongly associated with “excellent service” than with “poor service”. Negative difference scores indicated the reverse. Difference scores greater than zero were recoded as +1 and difference scores less than zero were recoded as −1 (scores at zero were not included in the subsequent analyses). For each attribute, we then computed the percentage number of 1 s, this value would reflect the percentage of participants who more strongly associated the prime with excellent service than with poor service.

3.2.5. Results

The results focus on the comparison of emotional attributes that were both significantly associated with providing versus receiving positive and negative customer service, as well as the overall number of positive and negative emotions attributed to each condition. Here, we first report the results for the entire group (averaged across all countries).

Providing Excellent Customer Service

The analysis of reaction times recorded when participants classified the targets subsequent to emotional primes in Test A (Providing customer service) found that (collapsed across all countries) providing excellent service was associated with fastest responses to feeling “calm” and “proud” ( p < 0.001). Other emotional word attributes that were found to be strongly associated with providing excellent service included feeling “fair”, “engaged”, “loved” and “pleased”, “nice”, “okay” and “ecstatic” ( p < 0.05). A total of nine positive emotion words were found to be implicitly associated with providing excellent customer service.

Receiving Excellent Customer Service

Receiving excellent service (Test B) was found to be associated with faster responses when preceded by the primes “energised”, “happy” and “proud” ( p < 0.001). Other emotions that were also strongly associated with receiving excellent customer service were “calm”, “satisfactory”, “nice”, “fair” and “okay” ( p < 0.05), attributes that were previously categorised as being experienced when engaging in everyday pleasures, such as meeting friends.

A comparison of significant associations across the two tests (see also Figure 2 ) revealed that only providing excellent service was associated with “pleased” and “ecstatic”, whereas receiving excellent service elicited significant associations with the attributes “energised”, “happy”, “thrilled”, “excited”, “fine” and “fortunate”.

Emotional attributes associated with providing versus receiving excellent customer service (Y-axis shows the percentage of people significantly associating primes with the receipt and provision of excellent service).

Providing Poor Customer Service

Providing poor customer service was significantly associated with the emotional attributes, “lonely”, “nervous”, “sad” and “annoyed”.

Receiving Poor Customer Service

Receiving poor customer service was significantly associated with these same four emotional attributes and additionally, with feeling “ignored”. Receiving poor service was more strongly associated with the attributes “sad” and “annoyed” than was providing poor service (see also Figure 3 ).

Emotional attributes associated with providing versus receiving poor customer service (Y-axis shows the percentage of people significantly associating primes with the receipt and provision of poor service).

Gender Differences

The statistical comparison of males and females collapsed across all countries found that while females associated more positive attributes with receiving excellent customer service, males associated more positive attributes with providing excellent customer service (both ps < 0.01). Specifically, females associated receiving excellent service with feeling “happy”, “energised”, “over-joyed”, “proud”, “thrilled”, “exhilarated”, “loved”, “nice”, “expected”, and “fair”. Receiving poor service was more significantly associated with feeling “nervous”, “sad” and “lonely”. By comparison, providing excellent service was associated with feeling “proud”, “calm”, “pleased”, “fair”, and “engaged”, whereas providing poor service was associated with feel “sad”, “annoyed” and “lonely”.

Males were faster to associate the provision of positive customer service with feeling “energised”, “calm”, “engaged”, “proud”, “thrilled” and “nice”. Providing poor service made them feel “nervous”. Receiving excellent customer service was associated with feeling “satisfactory”, “calm”, “okay”, “satisfied”, “ecstatic”, ”engaged”, “relief” and “nice”. Receiving poor customer service was associated with feeling “annoyed”, “regular” and “lonely”.

Age Differences

Respondents aged between 18 and 35 years associated more positive attributes with receiving than providing excellent customer service ( p < 0.001), including “OK, fair, confident, nice, satisfactory, engaged, energised, thrilled, calm, ecstatic, exhilarated, content, happy, pleasant”. Receiving poor service was associated with feeling “annoyed”, “sad” and “lonely”. Providing excellent service was associated with feeling “calm”, “engaged” and “proud” and “OK”; providing poor service made them feel “sad”, “ignored” and “nervous”.

In stark contrast, respondents in the older age group (36+) associated more positive attributes with providing rather than receiving excellent customer service ( p < 0.001). Specifically, providing excellent service was more closely associated with feeling “proud” and “calm”, “excited”, “pleased”, “nice” and “fair”. Providing poor service made them feel more “annoyed” and “sad”. Receiving excellent service made the older group feel “proud” and “calm”, poor service interactions made them feel “nervous”, “lonely”, “ignored”, “regular” and “sad”.

Cross-Cultural Differences

There were also a number of interesting cross-cultural differences in terms of the emotions most closely associated with providing and receiving good and poor customer service.

United Kingdom (least Impacted by Customer Service- Expectations much Lower)

Comparison of the statistical effect sizes between countries revealed that while respondents in the United Kingdom showed a positive association between giving or receiving amazing service, the effect was lower than that recorded for Canada and Australia.

Canada (Focused on “Providing”)

It was noteworthy that Canadians felt more “thrilled”, “content” and “pleased” when providing rather than receiving excellent service. Receiving, rather than giving amazing service was, on the other hand, more associated with a positive association with the emotions, “exhilarated”, “energised”, “happy”, “loved”, “relieved”, “pleasant” and “fine” ( p < 0.05).

Australia (Receiving is More Emotionally Important than Giving)

Australians were found to associate the provision of amazing service with a sense of “calm” ( p < 0.05), compared to receiving the same level of service. Australians were statistically more likely to feel “fortunate”, “thrilled”, “happy” and “appreciated” when receiving excellent service compared to when they were providing it.

3.2.6. Conclusions

In study 2, we demonstrated the implicit association of positive and negative feelings with proving and receiving good customer service across a large general populace. We also show the generalizability of our results across three cultures, ages and genders. Specifically, we demonstrated that, (1) providing and receiving excellent customer service was strongly associated with certain emotions (feeling “calm”, “proud”, “fair”, “engaged”, “loved”, “pleased”, “nice”, “okay”, “ecstatic”, “energised”, “happy” and “satisfactory”), and (2), providing and receiving poor customer service was strongly associated with certain emotions (feeling “lonely”, “nervous”, “sad”, “annoyed” and “ignored”), (3) females associated providing and receiving excellent customer service with certain emotions (“happy”, “energised”, “over-joyed”, “proud”, “thrilled”, “exhilarated”, “loved”, “nice”, “expected”, “fair”, “calm”, “pleased” and “engaged”), (4) females associated providing and receiving poor customer service with the emotions “nervous”, “sad”, “lonely” and “annoyed”, (5) males associated providing and receiving excellent customer service with the emotions “energised”, “calm”, “engaged”, “proud”, “thrilled”, “nice”, “satisfactory”, “okay”, “satisfied”, “ecstatic” and “relief”, (6) males associated providing and receiving poor customer service with the emotions (“nervous”, “annoyed”, “regular” and “lonely”). We also found that younger respondents associated more positive attributes with receiving, rather than providing, excellent customer service, whereas older respondents associated more positive attributes with providing rather than receiving excellent customer service. Among cross-cultural differences, we found that in (1), UK respondents showed a weak association between giving or receiving an amazing service and their expectations were lower (compared to Canada and Australia), (2) Canadian respondents showed a stronger association for providing rather than receiving excellent service and (3), Australian respondents showed a stronger association for receiving rather than providing excellent service).

4. General Discussion

In the current study, we exploited two implicit reaction time tasks. The first, the recently developed Impulse test, is a novel implicit reaction time paradigm that measures the moment-to-moment shifts in emotions when, for example, people are viewing dynamic videos or footages [ 34 ]. The second is a task based on affective priming, a very well established implicit paradigm that was developed out of cognitive psychology in the 1980s [ 41 , 42 , 43 ] and has been recently adapted for use in commercial neuromarketing studies [ 35 ]. Both implicit tasks are ideal for capturing the complex, often subconscious, emotions associated with receiving and providing customer service of varying quality in order to understand the subtle impact of these customer–staff interactions on emotional well-being. Major advantages of using these methods are that they are indirect and are not as susceptible to the response biases associated with explicit responses (e.g., self-reported measures) and that they can reveal the moment-to-moment scores during a video clip, rather than a post test score.

Our results show that people do not only find receiving excellent customer service as pleasurable but providing excellent service is equally satisfying. We corroborate these results using both physiological measures (study 1) and an implicit reaction time paradigm (study 2). We also provide evidence that both giving and receiving excellent service can actually reduce stress and anxiety levels amongst both consumers and service providers and have a positive impact on their wellbeing. These results were shown to hold true across three countries, demonstrating that giving and receiving excellent customer service can induce a sense of pride, calmness and of being loved.

Our data additionally revealed some age and gender differences. Specifically, our results reveal that younger individuals (18–35 years) exhibit more positive emotions when receiving than giving good customer service, whilst the opposite was the case for older participants. In thinking about being served, relatively more focus is placed on oneself (vs. others); in thinking about providing service, relatively more focus is placed on others (vs. oneself). Therefore, our results suggest that younger people tend to focus more on themselves (vs. others), whereas older individuals focus more on others (vs. themselves). This pattern of findings is consistent with Freund Blanchard-Fields’ [ 44 ] observation that older adults are more altruistic (i.e., focusing on the needs of others rather than on themselves) than younger adults, and tend to behave in ways that benefit others rather than themselves (e.g., donating money to a good cause rather than keeping it for themselves). By contrast, younger adults tend to focus on maximizing their personal gains over the interests of other people. Collectively, these findings add to the existing knowledge about customer service by underscoring the importance of age differences when it comes to customers and service providers. Future research may test the altruism explanation for the observed effects due to age differences.

Analysis of gender differences revealed that females tend to prefer receiving (vs. providing) excellent service, whereas the reverse is true for males. At first glance, this finding appears somewhat contradictory to past research that suggested that women are generally communal, warm, and nurturing, whereas men tend to be more competitive and goal-oriented [ 45 , 46 ]. However, we interpret this finding in the light of other research which showed that men and women place a different emphasis on different aspects of service. While men are usually more concerned about the core aspect of the service (e.g., the haircut received at a hair salon), women generally pay more attention to the relational aspects of service (e.g., how well one gets along with the hairstylist) [ 47 ]. It is also likely that the core (relational) aspects of service are more salient when thinking about giving (receiving) excellent customer service because the focus is on helping the recipient resolve their problem (core aspect); in thinking about receiving service, it is easier to think about how one would feel about being served (relational aspect). Applying this to the gender differences that we found, it is possible that men preferred giving (vs. receiving) excellent service because it is more closely aligned with their goal-oriented tendency. Women, on the other hand, preferred receiving (vs. giving) excellent service as they were drawn towards its more highly salient relational aspects of service as they imagine themselves being served. Future research may follow this lead to explicitly examine the underlying processes driving the results that we observed through the implicit tests.

4.1. Theoretical and Methodological Contributions

To the best of our knowledge, this research is the first to employ two implicit tests, targeting both individual and group level responses, in order to yield a comprehensive view of the payoffs of good customer service. This current research also contributes to retailing research, which tends to focus on explicit data, by adding the implicit angle to understand how customer service impacts individuals at a subconscious level.

4.2. Implications for Managers and Organizations

Past research on customer service is heavily focused on understanding how customer service affects the customer and how satisfied customers in turn reward organizations with increased sales, patronage, and higher profits. Service personnel, who often shoulder the “burden” of delivering customer service that yields benefits to customers and organizations, appear to gain the least from the exchange. Our current research augments this stream of literature by focusing on what customer service means to service providers. Managerially, the observation that older people exhibit a preference for providing good customer service suggests that companies might wish to consider employing more mature individuals on the front line (albeit with due consideration of the physical requirements related to standing in-stores for long hours) because they may be more naturally inclined to servicing the needs of others. In addition, we believe that our results gain credibility from the fact that for the implicit reaction time test, the primes chosen (e.g., pleasant experiences) were selected by real consumers and the targets (e.g., “being helpful”) were chosen in consultation with service industry consultants.

Our study found that service providers also benefit from delivering good customer service in the form of enhanced emotional well-being and inoculation against negative, damaging emotions. To some extent, understanding that delivering good customer service is emotionally lifting to the service providers helps to resolve the pressure of having to engage in acting to please customers. In the emotional labour literature, researchers identified two levels of acting—surface (where the employee displays false emotions that s/he does not feel, only to please customers) and deep (where the employee feels the emotions that she/he displays to customers)—that service personnel use when dealing with customers. However, both surface and deep acting have potential problems. Surface acting is often perceived as fake and distancing to customers; on the other hand, deep acting places considerable emotional strain on the service provider. Based on our results, service providers can be coached to focus on understanding how delivering good service makes them feel and the subsequent emotional payoffs they can gain from it. This may help to reduce employee burn-out and turnover whilst maintaining happy customers and a healthy bottom-line. Therefore, training employees to focus on how good customer service benefits themselves creates a positive feedback loop that benefits customers, service providers, and organizations alike.

Author Contributions

G.A.C.: Design, interpretation and co-drafting of the original manuscript; E.P.F.: Design and co-drafting of original manuscript; G.T.: Design, testing and analysis; A.P.: Literature review, interpretation and editing of manuscript; L.E.A.C.: Literature review, interpretation and editing of manuscript.

This research received no external funding.

Conflicts of Interest

The authors declare no conflict of interest.

Methods of Customer Service Research

Customer service research is done by organizations to measure how satisfied the consumer is with their brand’s services and products. Every brand aims to improve its customer experience and part of doing this is to measure, observe and check online surveys, personal messages, and more to ensure that their customer service performance is up to the mark.  According to experts, several steps can be researched and used to make sure that businesses’ customer service methods are high. Here are some of them.

• When looking at customer service, if you want to get the best out of it, consider separating it into three ways. This can be email inquiries, complaints, or issues that need to be transferred to tech support and regular phone calls with customer reps. This will help you to do a proper survey and help improve your customer service department. 
• If you are a project manager or a marketing research leader then one other thing you can do is to meet with the customer service director or manager and ask them the kind of questions they require. Suggest questions that can bring together timeliness, professionalism, wit, and accuracy in handling calls from customers. 
• You can even find a list of the various areas that customer service people cover, which can be the emails they use, the customers who phone in, and the ones that require to be transferred to other departments for customer support. This list can be gotten from the customer service department as they will normally keep records of these calls and messages. 
• Consider who will conduct telephone surveys, will this be asked to customers at the beginning or end of each call? Will it be conducted by temporary staff or full-time employees? Once you find out a preferred source for a marketing research agency then you can accordingly hire the right kind of employees to make the call. 
• Talk to the people conducting the survey and instruct them to pick one person randomly from the list that they call from. Ensure that every third or second person helps eliminate the bias from the survey and the call. This can help you pick between three groups of people.

Final verdict

When it comes to methods or ways to improve customer service experience, ensure that you take a customer service survey by using an online website calculator. Take each list including phone calls, emails, technical support, and more on a separate list and note down the common percent used. This can help you enter the complete number of names you require for the customer service groups. Customer service is a very important factor for every brand and organization and you need to ensure that when your customers call their issues and problems are being handled efficiently and professionally. Taking a survey can help you do the same and can also help you to determine which are doing well and working to the benefit of your business and customers and which area needs to improve and better their game. 

About the Author

Himanshu vashishtha administrator.

Find A Dealer

Order Status

Chat with an Expert

877-216-5446

Talk to an Expert

All Products

Dealer Support

Contact AMP Research

Customer Service Hours

Monday - Friday 7AM-5PM PT

• Side Hustles
• Power Players
• Young Success
• Save and Invest
• Become Debt-Free
• Land the Job
• Closing the Gap
• Science of Success
• Pop Culture and Media
• Psychology and Relationships
• Health and Wellness
• Real Estate
• Most Popular

Related Stories

• Work The majority of Americans are not   engaged at work—here's why
• Work Gen AI is here to stay – here's how to stay   relevant in the changing job market
• Next Gen Investing ChatGPT can help people in these 3   industries, says OpenAI CEO Sam Altman
• Work U.S. company tested a 4-day workweek—says   workers are happier, more productive
• Get Ahead 5 high-paying tech skills   companies are hiring for now

Stanford and MIT study: A.I. boosted worker productivity by 14%—those who use it 'will replace those who don't'

Artificial intelligence tools like chatbots helped boost worker productivity at one tech company by 14%, according to new research from Stanford and MIT that was first reported by Bloomberg .

The study is thought to be the first major real-world application of generative AI in the workplace. Researchers measured productivity of more than 5,000 customer support agents, based primarily in the Philippines, at a Fortune 500 enterprise software firm over the course of a year.

Tech support agents who used AI tools that created conversational scripts boosted their productivity, measured as issues resolved per hour, by 14% on average, but the improvement was even more pronounced for "novice and low-skilled workers" who were able to get their work done 35% faster.

In some cases, using AI trumped having real-life work experience: Customer service agents with two months of experience who used AI support performed as well or better than agents with over six months of experience working without AI.

Meanwhile, the use of AI tools showed a minimal impact on "experienced or highly skilled workers," the authors note, and at times served as a distraction.

AI support can be especially helpful to entry-level or early-career workers, says Lindsey Raymond, an MIT Ph.D. candidate and co-author of the paper. Less experienced workers benefit from AI by taking its recommendations to get up to speed and learn skillsets that usually come with experience.

With that said, AI tools benefit from the best and brightest workers training the AI itself by providing examples of best practices, which the technology then turns into recommendations for others workers to apply.

Businesses should understand that, despite less dramatic changes in productivity, high-performing employees should be recognized and compensated for generating the solutions that others can learn from, Raymond says.

Happier workers and customers

The year-long experiment also revealed that AI assistance improved customer satisfaction, reduced requests for managerial intervention and improved employee retention.

The research isn't meant to hypothesize whether AI will replace workers, Raymond says, but rather concludes the technology will help workers more effectively multitask and handle more complicated questions faster.

Better and faster work led to happier customers, who were in turn nicer to customer service agents and improved employee retention, Raymond says.

Tools that make people more effective at their jobs make the experience of work less stressful, she adds.

Results that generative AI can boost productivity is generally good news, though the biggest benefits may not be evenly distributed, says Erik Brynjolfsson, the director of the Digital Economy Lab at the Stanford Institute for Human-Centered AI, and co-author on the report. "There's no guarantee we'll all benefit, but it certainly sets the table for us all being better off," he says.

'Workers who work with generative AI will replace those who don't'

Brynjolfsson says call centers are a great place to use generative AI because it involves a lot of scripted language, but that "almost any kind of information or knowledge work that involves language could benefit from this," including across legal, marketing, medicine, teaching and other fields.

Workers at all levels can benefit from the technology, he adds — he recently spoke with a CEO who used generative AI to prepare for a board meeting.

"Probably over half of the U.S. workforce will be significantly affected by these tools," Brynjolfsson says.

He adds that workers, especially young workers, can stay ahead of the curve by embracing the reality of the technology: "Workers who embrace the technology, play around with it and learn how to use it are the ones that are going to succeed and benefit the most," Brynjolfsson says. "I don't think the generative AI is going to replace workers, but workers who work with generative AI will replace those who don't."

Some experts say generative AI tools could affect how two-thirds of current jobs are performed and could eventually raise global gross domestic product by as much as 7%, according to one economic report from Goldman Sachs .

One recent survey of LinkedIn's Top Companies found that nearly 70% say AI is already helping them be faster and smarter, and another 32% say they expect to see larger gains from using AI in the coming years. And companies like EY explicitly listed AI as one of their top three hiring priorities, while Wells Fargo and Kaiser Permanente are implementing AI across their workflows .

Want to be smarter and more successful with your money, work & life?  Sign up for our new newsletter !

Check out: Meta has a new AI tool to fight misinformation—and it’s using Wikipedia to train itself

We've detected unusual activity from your computer network

To continue, please click the box below to let us know you're not a robot.

Why did this happen?

Please make sure your browser supports JavaScript and cookies and that you are not blocking them from loading. For more information you can review our Terms of Service and Cookie Policy .

For inquiries related to this message please contact our support team and provide the reference ID below.

LLM2Vec: Large language models are secretly powerful text encoders

• ServiceNow Research

Authors: Parishad BehnamGhader, Vaibhav Adlakha, Marius Mosbach, Dzmitry Bahdanau, Nicolas Chapados, Siva Reddy

Text-embedding models convert a piece of text, such as a search query, document, or piece of code, into a sequence of real-valued numbers. Given such embeddings, we can measure the similarity, or relatedness, of pieces of text. This facilitates various important applications, such as search, clustering, retrieval, and classification.

With the widespread availability of decoder-only large language models (LLMs), such as GPT-4, LLaMA2, Mistral-7B, and StarCoder2 , a pressing question in the natural language processing (NLP) research community is how best to use these models to construct powerful text embeddings.

We’re excited to present LLM2Vec: Large Language Models Are Secretly Powerful Text Encoders , a simple and efficient solution to transform any decoder-only LLM into a powerful text encoder in an unsupervised fashion simply by using adapters (LoRA), without the need to modify the base models.

Below we give an overview of the key components of LLM2Vec and present the exciting results we got when benchmarking LLM2Vec models on the challenging Massive Text Embeddings Benchmark (MTEB) . Our LLM2Vec-Mistral ranks first on the MTEB leaderboard in the unsupervised category, first in the supervised category among the models trained on publicly available embedding data (E5), and seventh on the overall leaderboard (the other top six models are trained on synthetic data generated from GPT-4/similar-scale models).  

A simple and efficient recipe

At its core, LLM2Vec consists of three simple steps:

Enabling bidirectional attention

Adaptation via masked next-token prediction (MNTP)

Adaptation via unsupervised contrastive learning

Adapting a model with the LLM2Vec approach is highly efficient and works with parameter-efficient fine-tuning methods such as LoRA. Additionally, the adaptation can be performed using a general domain corpus such as Wikipedia, requires only a few hundred training steps, and can be run on a single GPU.

State-of-the-art performance

LLM2Vec is not only simple and efficient, but it also leads to state-of-the-art performance on the challenging MTEB, both in the unsupervised and supervised setting (among models trained only on publicly available data).  

Unsupervised results

We applied LLM2Vec to some of the best-performing LLMs available and evaluated the resulting text--embedding models on MTEB. In the unsupervised setting—i.e., without using any labeled training data for contrastive learning—our LLM2Vec-transformed models achieved a new state-of-the-art performance of 56.80, outperforming the previous unsupervised approach by a large margin.

Supervised results

LLM2Vec can also be easily combined with supervised contrastive learning. As our results show, applying LLM2Vec before supervised contrastive learning leads to a substantial improvement. Moreover, LLM2Vec in combination with Mistral-7B, currently the best-performing 7 billion-parameter LLM, leads to a new state-of-the-art performance of 64.80 on MTEB among models trained only with publicly available data.

Highly sample-efficient

LLM2Vec-transformed models require less training data to perform well compared to training models without the LLM2vec transformation.  

These results make us particularly excited about challenging real-world scenarios where large amounts of labeled data might be costly to acquire.  

Use it on your own data 

We’ve made it easy for you to use our LLM2Vec-transformed models. LLM2Vec class is a wrapper on top of Hugging Face models to support sequence encoding and pooling operations. The steps below showcase an example of how to use the library.  

Preparing the model 

Here, we first initialize the model and apply MNTP-trained LoRA weights on top. After merging the model with MNTP weights, we can either:

Load the unsupervised-trained LoRA weights (trained with SimCSE objective and wiki corpus) 

Load the model with supervised-trained LoRA weights (trained with contrastive learning and public E5 data)  

Applying LLM2Vec wrapper

Then, we define our LLM2Vec encoder model as follows:

This model now returns the text embedding for any input in the form of [[instruction1, text1], [instruction2, text2]] or [text1, text2]. While training, we provide instructions for both sentences in symmetric tasks and only for queries in asymmetric tasks.  

As demonstrated above, LLM2Vec is a simple unsupervised approach that can transform any pretrained decoder-only LLM into a strong text encoder. If you’re as excited about LLM2Vec as we are, check out our hands-on tutorial , which walks you through the different steps of our method. We also welcome contributions on Github and invite the community to share their LLM2Vec-transformed models.

Research:  Project page

Code:  LLM2Vec on GitHub

Tutorial:  Learn how to apply LLM2Vec to LLaMA-2

Find out more about ServiceNow Research .  

© 2024 ServiceNow, Inc. All rights reserved. ServiceNow, the ServiceNow logo, Now, and other ServiceNow marks are trademarks and/or registered trademarks of ServiceNow, Inc. in the United States and/or other countries. Other company names, product names, and logos may be trademarks of the respective companies with which they are associated.

Introducing the ServiceNow Scholarship to promote diversity in tech

Meet the 2023 ServiceNow Worldwide Partner Award winners

When great customer service meets AI (in Australia!)

Trends & research.

ServiceNow is a Leader in third-party risk management (TPRM) platforms

IDC MarketScape names ServiceNow a Leader in SPM and value stream management

A Magic Quadrant Leader in ITSM Platforms for ninth year in a row

• United States - Global

Asia, Pacific, and Japan

Europe, Middle East, and Africa

• United Kingdom - English
• Deutschland - Deutsch
• France - Français
• Nederland - Nederlands

An official website of the United States government

Here’s how you know

Official websites use .gov A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

• Explore sell to government
• Ways you can sell to government
• How to access contract opportunities
• Conduct market research
• Register your business
• Certify as a small business
• Become a schedule holder
• Market your business
• Research active solicitations
• Respond to a solicitation
• What to expect during the award process
• Comply with contractual requirements
• Handle contract modifications
• Monitor past performance evaluations
• Explore real estate
• 3D-4D building information modeling
• Art in architecture | Fine arts
• Computer-aided design standards
• Commissioning
• Design excellence
• Engineering
• Project management information system
• Spatial data management
• Facilities operations
• Smart buildings
• Tenant services
• Utility services
• Water quality management
• Explore historic buildings
• Heritage tourism
• Historic preservation policy, tools and resources
• Historic building stewardship
• Videos, pictures, posters and more
• NEPA implementation
• Courthouse program
• Land ports of entry
• Prospectus library
• Regional buildings
• Renting property
• Visiting public buildings
• Real property disposal
• Reimbursable services (RWA)
• Rental policy and procedures
• Site selection and relocation
• For businesses seeking opportunities
• For federal customers
• For workers in federal buildings
• Explore policy and regulations
• Acquisition management policy
• Aviation management policy
• Information technology policy
• Real property management policy
• Relocation management policy
• Travel management policy
• Vehicle management policy
• Federal acquisition regulations
• Federal management regulations
• Federal travel regulations
• GSA acquisition manual
• Managing the federal rulemaking process
• Explore small business
• Explore business models
• Research the federal market
• Forecast of contracting opportunities
• Events and contacts
• Explore travel
• Per diem rates
• Transportation (airfare rates, POV rates, etc.)
• State tax exemption
• Travel charge card
• Conferences and meetings
• E-gov travel service (ETS)
• Travel category schedule
• Federal travel regulation
• Travel policy
• Explore technology
• Cloud computing services
• Cybersecurity products and services
• Data center services
• Hardware products and services
• Professional IT services
• Software products and services
• Telecommunications and network services
• Work with small businesses
• Governmentwide acquisition contracts
• MAS information technology
• Software purchase agreements
• Cybersecurity
• Digital strategy
• Emerging citizen technology
• Federal identity, credentials, and access management
• Mobile government
• Technology modernization fund
• Explore about us
• Annual reports
• Mission and strategic goals
• Role in presidential transitions
• Get an internship
• Launch your career
• Elevate your professional career
• Discover special hiring paths
• Events and training
• Agency blog
• Congressional testimony
• GSA does that podcast
• News releases
• Leadership directory
• Staff directory
• Office of the administrator
• Federal Acquisition Service
• Public Buildings Service
• Staff offices
• Board of Contract Appeals
• Office of Inspector General
• Region 1 | New England
• Region 2 | Northeast and Caribbean
• Region 3 | Mid-Atlantic
• Region 4 | Southeast Sunbelt
• Region 5 | Great Lakes
• Region 6 | Heartland
• Region 7 | Greater Southwest
• Region 8 | Rocky Mountain
• Region 9 | Pacific Rim
• Region 10 | Northwest/Arctic
• Region 11 | National Capital Region
• Per Diem Lookup

Market Research as a Service

A value-added service at no additional cost.

MRAS delivers meaningful market data to federal, state, and local agencies. Through FAR Part 10 compliant requests for information, sources sought, industry days, and advanced product research, MRAS collects data to help customers understand where their need fits within the GSA government-wide marketplace.

Watch a video to learn more about Market Research As a Service .

Explore our offerings

MRAS provides automated RFIs and Sources Sought for services and advanced GSA Advantage product searches.

• Request for Information (RFI) with Market Research Report  — Request an RFI to understand GSA contracts and industry capabilities with a market report.
• Product Research Request  — Search up to 20,000 products on GSA Advantage and receive a market report.
• MRAS Report Archives - COMING SOON! Search thousands of previous market research reports to research your requirement (OMB Max Login required).

Request for information with market research report

Develop customized RFIs that strategically target GSA contract holders on GSA eBuy, resulting in a robust market research analysis report showing all responses received, in as little as 1-2 weeks, depending on the requirement. Within 24 hours of receiving your request, we work directly with agency POCs to review a draft RFI developed from information provided, and work side-by-side with customers to further refine the RFI until it is ready to be posted on GSA eBuy.

After the RFI closes, a Market Research Report is designed specifically to see how the GSA market can meet technical requirements and mission needs provided. Market Research Reports are designed to facilitate completing Small Business Office review forms, developing Acquisition Plans, and assisting in formulating an overall acquisition strategy.

Begin your RFI or product market research request .

What to expect in the MRAS process

• Customer submits a request for an RFI.
• Collaborate with us and develop RFI.
• Customer approves and engages industry.
• RFI closes and you get a market report.
• Consult with us.

The total estimated time is one day to two weeks, depending on the requirement.

MRAS provides each customer with a unique, comprehensive, and easy-to-understand report. The report analyzes industry partner data related to the customers' needs, which gives an understanding of the results expected under the markets researched. The Market Research Report includes socio-economic, technical, capabilities, and comprehensive business information so agencies can complete their Acquisition Planning and Small Business strategies and requirements and estimate documents.

Training and events

Mras customer training on effective market research (for federal, military, state, and local).

Attention all public sector employees in the contracting and acquisition career field! In GSA's Market Research As a Service Customer Training: Effective Market Research (Fed/Military), you'll learn the importance of market research, how and when to conduct it, and how to get the best results by making your data collection methods more efficient.

You'll also delve into the regulatory nature of FAR Part 10 and how it ties into other decisions such as acquisition planning, small business set-asides, commercial items, and contract type selection.

This course includes real-life scenarios, expert instructors, and tools from GSA's market research as a service. Take this opportunity to enhance your market research skills and advance your career. This is a one-CLP credit course.

Register Today

MRAS industry training

As an industry partner, you can provide valuable market research data to inform the purchasing decisions of GSA customers. Participating in our research allows you to showcase your products and services to potential buyers and play a key role in helping government agencies make informed and efficient purchasing decisions.

But why choose our MRAS over traditional market research methods? Our research is 100 times easier and efficient, making it a great business development activity for your company. Plus, by participating in MRAS, you can reach a wider audience of government buyers and increase your visibility in the market.

We offer monthly industry training seminars to help you get started and make the most of your participation in MRAS. CLPs are not issued for these webinars.

MRAS in the news

• Listen to the FAS Focus: Deep Dive podcast episode as we discuss GSA’s Market Research As a Service with Kevan Kivlan, Director, Customer & Stakeholder Engagements New England & FAS MRAS Program Office: https://twitter.com/i/status/150-165-6591321518081
• January 2022: MRAS Pushes Billions in Business to the GSA Schedule
• June 2021: MRAS Webinars Show Industry How To Work Smarter Not Harder

What our customers say

“Utilizing this resource is a great way to leverage the hard work put into establishing these contracts and provide a great benefit to the Government as a whole by simplifying the already complex process of federal acquisition. The benefit of conducting this market research gives a PCO the ability to see if there's a more expedient or cost-effective way to support their mission partner's needs. Often contracting is known as a slow and drawn-out process, and while it can be, this free tool is fast and gives a great snapshot as to what alternatives are available. MRAS is a great tool in a PCO's toolkit to have and one that should not be overlooked.”

— Capt. Jeremy A. Deorsey, Massachusetts Army National Guard

MRAS background

What MRAS is: MRAS helps customers by collecting and providing industry partner information specific to Agency needs. MRAS conducts Requests for Information, Sources Sought, and advanced product searches to help Agencies compile business, socio-economic, and technical data. Additionally, MRAS helps provide feedback from industry about Agency requirements documents, industry trends, best practices, and the best GSA contract and NAICs for their acquisition.

Cost: MRAS is a market research service GSA provides to all federal, state, and local agencies at no cost.

How to submit your market research request: Please reach out to your agency POC  so they can walk you through the MRAS process and determine the best MRAS service for your requirement. You can then fill out the MRAS Service Request Form .

For industry partners that need help with a survey

Please fill out the Industry Help Request Form .

Still have questions?

Contact your agency POC or email the MRAS team at [email protected] .

PER DIEM LOOK-UP

1 choose a location.

Error, The Per Diem API is not responding. Please try again later.

No results could be found for the location you've entered.

Rates for Alaska, Hawaii, U.S. Territories and Possessions are set by the Department of Defense .

Rates for foreign countries are set by the State Department .

2 Choose a date

Rates are available between 10/1/2021 and 09/30/2024.

The End Date of your trip can not occur before the Start Date.

Traveler reimbursement is based on the location of the work activities and not the accommodations, unless lodging is not available at the work activity, then the agency may authorize the rate where lodging is obtained.

Unless otherwise specified, the per diem locality is defined as "all locations within, or entirely surrounded by, the corporate limits of the key city, including independent entities located within those boundaries."

Per diem localities with county definitions shall include "all locations within, or entirely surrounded by, the corporate limits of the key city as well as the boundaries of the listed counties, including independent entities located within the boundaries of the key city and the listed counties (unless otherwise listed separately)."

When a military installation or Government - related facility(whether or not specifically named) is located partially within more than one city or county boundary, the applicable per diem rate for the entire installation or facility is the higher of the rates which apply to the cities and / or counties, even though part(s) of such activities may be located outside the defined per diem locality.

Wiz Research finds architecture risks that may compromise AI-as-a-Service providers and consequently risk customer data; works with Hugging Face on mitigations

Wiz researchers discovered architecture risks that may compromise AI-as-a-Service providers and put customer data at risk. Wiz and Hugging Face worked together to mitigate the issue.

The world has never seen a piece of technology adopted at the pace of AI. As more organizations worldwide adopt AI-as-a-Service (a.k.a. “AI cloud”) the industry must recognize the possible risks in this shared infrastructure that holds sensitive data and enforce mature regulation and security practices that are similar to those enforced on public cloud service providers.  

When we move fast, we break things. In recent months, Wiz Research partnered with AI-as-a-Service companies to uncover common security risks that may impact the industry and subsequently put users’ data and models at risk. In our State of AI in the Cloud report , we show that AI services are already present in more than 70% of cloud environments, showcasing how critical the impact of those findings are. 

In this blog we outline our joint work with Hugging Face, one of the best-known AI-as-a-Service providers. Hugging Face has undergone a meteoric rise and grown at an unprecedented rate to meet swelling demand. What we found not only presented an opportunity for Hugging Face to strengthen the platform’s security (which they did); it also carries broader takeaways that apply to many AI systems and AI as-a-service platforms. 

AI models require strong GPU to run, which is often outsourced to AI service providers. In Hugging Face, this service is called Hugging Face Inference API. For ease of understanding, this can be compared, at a high level, to consuming cloud infrastructure from AWS/GCP/Azure to run your applications and code. Wiz Research was able to compromise the service running the custom models by uploading our own malicious model and leveraging container escape techniques to break out from our tenant and compromise the entire service. This means Wiz research could gain cross-tenant access to other customers' models stored and run in Hugging Face.  

We believe those findings are not unique to Hugging Face and represent challenges of tenant separation that many AI-as-a-Service companies will face, considering the model in which they run customer code and handle large amounts of data while growing faster than any industry before. We in the security community should partner closely with those companies to ensure safe infrastructure and guardrails are put in place without hindering this rapid (and truly incredible) growth. 

We want to thank the Hugging Face team for their collaboration and partnership. They have published their own blog post in response to our research, detailing the events and outcomes from their perspective.  

About Hugging Face 

Hugging Face stands out as the de facto open and collaborative platform for AI builders with a mission to democratize good Machine Learning. It provides users with the necessary infrastructure to host, train, and collaborate on AI model development within their teams. In addition to these capabilities, Hugging Face also serves as one of the most popular hubs where users can explore and utilize AI models developed by the AI community, discover and employ datasets, and experiment with demos. As part of its mission, Hugging Face feels a responsibility to keep up to date with AI/ML risks .  

Being a pivotal player in the broader AI development ecosystem, Hugging Face has also become an attractive target for adversaries. If a malicious actor were to compromise Hugging Face's platform, they could potentially gain access to private AI models, datasets, and critical applications, leading to widespread damage and potential supply chain risk. 

What did we find?   

Malicious models represent a major risk to AI systems, especially for AI-as-a-service providers because potential attackers may leverage these models to perform cross-tenant attacks. The potential impact is devastating, as attackers may be able to access the millions of private AI models and apps stored within AI-as-a-service providers. Wiz found two critical risks present in Hugging Face’s environment that we could have taken advantage of:  

Shared Inference infrastructure takeover risk – AI Inference is the process of using an already-trained model to generate predictions for a given input. Our research found that inference infrastructure often runs untrusted, potentially malicious models that use the “pickle” format. A malicious pickle-serialized model could contain a remote code execution payload, potentially granting the attacker escalated privileges and cross-tenant access to other customers' models.  

Shared CI/CD takeover risk – compiling malicious AI applications also represents a major risk as attackers may attempt to take over the CI/CD pipeline itself and perform a supply chain attack. A malicious AI app could have done so after taking over the CI/CD cluster.  

Different types of AI/ML applications 

When thinking about security for AI/ML, it is important to distinguish between different types of applications and scopes. An average application that uses AI/ML would consist of the following components: 

Model : The AI models that are being used (i.e. LLaMA, Bert, Whisper, etc.). 

Application : The application code that passes inputs to the AI model and makes use of the predictions it creates. 

Inference Infrastructure : The infrastructure that allows execution of the AI model — being “on edge” (like Transformers.js ) or via API or Inference-as-a-Service (like Hugging Face’s Inference Endpoints ). 

Potential adversaries can choose to attack each of the above components via different methods. For instance, to attack the AI model specifically, attackers can use certain inputs that would cause the model to produce false predictions (like adversarial.js ). To attack the application that utilizes AI/ML, attackers can use an input that produces a prediction that is correct — but is being used unsafely within the application (for instance, producing a prediction that would cause an SQL injection to the database, since the application would consider the output prediction of the model to be a trusted input).  

Finally, it is also possible to attack the inference infrastructure by utilizing a specially crafted, pickle-serialized malicious model. It is very common to treat AI models as black-box and to utilize other publicly available AI models. Currently, there are very few tools that can be used to examine the integrity of a given model and verify that it is indeed not malicious (such as Pickle Scanning by Hugging Face) — so developers and engineers must be very careful deciding where to download the models from. Using an untrusted AI model could introduce integrity and security risks to your application and is equivalent to including untrusted code within your application. 

In this blog post, we will demonstrate how to gain access to Hugging Face’s infrastructure with a special handcrafted serialization exploit, and detail what can be done to minimize the risk. 

The AI security questions and findings 

The Wiz research team is highly focused on isolation vulnerabilities in cloud environments. When we saw the rise of AI-as-a-service companies, we were concerned about the potential implications of a malicious actor leveraging them to gain privileged cross-tenant access, since AI models in practice are actually code. By design, AI-as-a-service providers build a shared compute service for their customers, which triggers an immediate question: is the AI model running in an isolated environment? Is it isolated enough?  

In this research, we focused on three key offerings of the platform: 

Inference API – which allows the community to browse and experiment with available models on the hub, without having to install required dependencies locally. Instead, users can interact with and “preview” these models via a modal on the platform, which is powered by Inference API. 

Inference Endpoints – which is a fully managed offering by Hugging Face that lets users easily deploy AI models on dedicated infrastructure for production purposes (i.e. Inference-as-a-Service). 

Spaces – which offers a simple way to host AI/ML applications, for the purpose of showcasing AI models or working collaboratively on developing AI-powered applications. 

Researching Hugging Face Inference API and inference endpoints 

When we, as attackers, examined the Inference offerings of Hugging Face (both Inference API and Inference Endpoints), we realized that any user could upload their own model. Behind the scenes, Hugging Face will dedicate resources, with the dependencies and infrastructure required for users to be able to interact with it and obtain predictions. 

This raised an interesting question: could we, as users of the platform, upload a specially crafted model – one could call it malicious – that would let us execute arbitrary code in that interface? And if we did manage to execute code inside Inference API, what would we find there? 

Uploading a Malicious Model to the Hub 

Hugging Face’s platform supports various AI model formats. By performing a quick search on Hugging Face, we can see that two formats are more prominent than others: PyTorch ( Pickle ) and Safetensors . 

It is relatively well-known that Python’s Pickle format is unsafe, and that it is possible to achieve remote code execution upon deserialization of untrusted data when using this format. This is even mentioned in Pickle’s own documentation :  

Because Pickle is an unsafe format, Hugging Face performs some analysis ( Pickle Scanning and Malware Scanning ) on Pickle files uploaded to their platform, and even highlights those they deem to be dangerous. 

Hugging Face will still let the user infer the uploaded Pickle-based model on the platform’s infrastructure, even when deemed dangerous. Because the community still uses PyTorch pickle, Hugging Face needs to support it. 

As researchers, we wanted to find out what would happen if we uploaded a malicious Pickle-based model to Hugging Face and interacted with it using the Inference API feature. Would our malicious code be executed? Would it run in a sandboxed environment? Do our models share the same infrastructure as those of other Hugging Face users? (In other words, is Inference API a multi-tenant service?) 

Let’s find out. 

Remote code execution via specially crafted Pickle file 

Without going into too much detail, we can state that it is relatively straightforward to craft a PyTorch (Pickle) model that will execute arbitrary code upon loading. To achieve remote code execution, we simply cloned a legitimate model ( gpt2 ), which already includes all of the necessary files that instruct Hugging Face on how this model should be run (i.e. config.json). We then modified it in a way that would run our reverse-shell upon loading. Next, we uploaded our hand-crafted model to Hugging Face as a private model and attempted to interact with it using the Inference API feature — and voila, we got our reverse shell! 

For convenience, instead of invoking a reverse shell every time we needed to check something, we chose to craft a version of our malicious model that could function like a shell. By hooking a couple functions in Hugging Face's python code, which manages the model's inference result (following the Pickle-deserialization remote code execution stage), we achieved shell-like functionality. The results are the following: 

Amazon EKS privilege escalation via IMDS 

After executing code inside Hugging Face Inference API and receiving our reverse shell, we started exploring the environment where we were running. It was quickly noticeable to us that we were running inside a Pod in a Kubernetes cluster hosted on Amazon EKS.  

In the past year, we encountered Amazon EKS multiple times during our research into service provider security issues. In fact, we have encountered Amazon EKS enough times to prompt us to create a playbook outlining what to look for when we see an EKS cluster (some of these key takeaways are documented in the 2023 Kubernetes Security report ). 

Following our “playbook” of common EKS misconfigurations (or insecure defaults) and how to identify them, we noticed that we could query the node’s IMDS (169.254.169.254) from within the pod where we were running. Since we could query the node’s IMDS and obtain its identity, we could also obtain the role of a Node inside the EKS cluster by abusing the aws eks get-token command . This is a fairly common misconfiguration (/ insecure default) in Amazon EKS. It is popular enough that we have included this exact trick in our EKS Cluster Games CTF (Challenge #4) even prior to doing this research. A small caveat with this method is that, in order to generate a valid token for the Kubernetes cluster, we must supply the correct cluster name to the aws eks get-token command. We tried guessing the correct cluster name a couple of times with no luck (based on the name of our AWS role), but eventually noticed that our AWS role also had permissions to call DescribeInstances (a default configuration), which revealed the name of the cluster via a tag attached to nodes’ compute. 

Using the aws eks get-token command and the IAM identity from the IMDS, we generated a valid Kubernetes token with the role of a Node. 

Now that we have the role of a node inside the Amazon EKS cluster, we have more privileges, and we can use them to explore the environment even further.  

One of the things we did was to list information about the Pod where we were running via kubectl get pods/$(hostname) , and then view the secrets that are associated with our pod. We were able to prove that by obtaining secrets (using kubectl get secrets ) it was possible to perform lateral movement within the EKS cluster.  

Potential impact and mitigations 

The secrets we obtained could have had a significant impact on the platform if they were in the hands of a malicious actor. Secrets within shared environments may often lead to cross-tenant access and sensitive data leakage.  

To mitigate this issue, we recommend enabling IMDSv2 with Hop Limit to prevent pods from accessing the IMDS and obtaining the role of a node within the cluster. 

Researching Hugging Face Spaces 

As we mentioned, Spaces is a different service in Hugging Face that allows users to host their AI-powered application on Hugging Face’s infrastructure for the purpose of collaborative development and showcasing the application to the public. Conveniently, all Hugging Face requires from the user in order to run their application on the Hugging Face Spaces service is a Dockerfile. 

Remote Code Execution via a specially crafted Dockerfile  

We began our engagement by providing a Dockerfile that executes a malicious payload via the CMD instruction, which specifies what program to run once the docker container is started. After gaining code execution and exploring the environment for a while, we found it to be quite restrictive and isolated. Subsequently, we decided to use the RUN instruction instead of the CMD instruction, enabling us to execute code in the build process and potentially encounter a different environment. 

Network isolation issue – write access to centralized container registry 

After executing code in the building process of our image, we used the netstat command to examine network connections made from our machine. One connection was to an internal container registry to which our built layers were pushed. This makes sense. An image was built, and it should be stored somewhere — this is a perfect application for a container registry. However, this container registry did not serve only us; it also served more of Hugging Face’s customers. Due to insufficient scoping, it was possible to pull and push (thus overwrite) all the images that were available on that container registry. 

In the wrong hands, the ability to write to the internal container registry could have significant implications for the platform's integrity and lead to supply chain attacks on customers’ spaces. To mitigate such issues, we recommend enforcing authentication even for internal container registries and, in general, limiting access to them. 

This research demonstrates that utilizing untrusted AI models (especially Pickle-based ones) could result in serious security consequences. Furthermore, if you intend to let users utilize untrusted AI models in your environment, it is extremely important to ensure that they are running in a sandboxed environment — since you could unknowingly be giving them the ability to execute arbitrary code on your infrastructure. The pace of AI adoption is unprecedented and enables great innovation. However, organizations should ensure that they have visibility and governance of the entire AI stack being used and carefully analyze all risks, including usage of malicious models, exposure of training data, sensitive data in training, vulnerabilities in AI SDKs, exposure of AI services, and other toxic risk combinations that may exploited by attackers. 

This research also highlights the value of collaboration between security researchers and platform developers. Collaboration of this type aids in gaining a deeper understanding of the risks associated with the platform, and ultimately enhances its security posture . 

Hugging Face has recently implemented Wiz CSPM and vulnerability scanning to proactively identify and mitigate some of the toxic risk combinations found here. In addition, Hugging Face is also currently going through its annual penetration test to ensure identified items have been sufficiently mitigated. 

Stay in touch!

Hi there! We are Sagi Tzadik ( @sagitz_ ), Shir Tamari ( @shirtamari ), Nir Ohfeld ( @nirohfeld ), Ronen Shustin ( @ronenshh ) and Hillai Ben-Sasson ( @hillai ) from the Wiz Research Team ( @wiz_io ). We are a group of veteran white-hat hackers with a single goal: to make the cloud a safer place for everyone. We primarily focus on finding new attack vectors in the cloud and uncovering isolation issues in cloud vendors and service providers. We would love to hear from you! Feel free to contact us on X (Twitter) or via email: [email protected] .

Continue reading

Defense in depth: XZ Utils

We explore assessment, prevention, and detection strategies for protecting your organization from the XZ Utils vulnerability.

Wizards of security, casting spells on themselves for ultimate digital security

Wiz practices what it preaches. Let’s look at how the security team at Wiz uses the power of the Wiz platform to monitor all its cloud-based infrastructure and services.  

Top security talks from KubeCon Europe 2024

KubeCon Europe is the largest open source community conference in Europe with hundreds of talks, many of them about security. All the sessions are available online; in this blog, we’ll discuss our favorites.

Ready to see Wiz in action?

“Best User Experience I have ever seen, provides full visibility to cloud workloads.”

“Wiz provides a single pane of glass to see what is going on in our cloud environments.”

“We know that if Wiz identifies something as critical, it actually is.”