A Beginner's Guide for Customer Experience Teams
Nov 03, 2026 - 13 minute read
A new customer experience team usually inherits a lot of signals and very little structure. There is a survey someone set up years ago, a support inbox, an app store page full of reviews and a social media account, and each is owned by a different team. The job of the CX team is to turn these into one practice: a clear set of metrics, a known list of listening channels, a regular routine for analyzing what customers say and a way of making sure the findings lead to change.
This guide describes what to set up first and why. It is written for people who have just joined or formed a CX team, whatever tools their company uses. The method for analyzing CX data in depth has its own guide, Customer Experience Analysis: A Step-by-Step Method, which this one links to where it matters.
What a CX Team Is Responsible For
A CX team rarely fixes problems itself. The checkout flow belongs to the product team, delivery belongs to operations and the tone of a campaign belongs to marketing. What the CX team owns is the understanding of the customer: listening to customers across the channels where they speak, measuring how the experience is going, explaining why it is going that way, bringing the evidence to the team that can act on it and checking afterwards whether customers noticed a difference.
This is broader than customer service. Customer service handles individual contacts, one ticket or one call at a time, and is one part of the experience. Customer experience is the perception a customer forms across the whole relationship, from first hearing about a product to buying it, using it, asking for help and deciding whether to stay.
The Three Core Metrics
Three survey metrics are standard in customer experience work. Each answers a different question, and each becomes far more useful when it is followed by one open-ended question asking why the customer gave that score.
Net Promoter Score (NPS)
The Net Promoter Score asks how likely the customer is to recommend the company to a friend or colleague, on a scale from 0 to 10. Respondents who answer 9 or 10 are Promoters, 7 or 8 are Passives and 0 to 6 are Detractors. NPS is the percentage of Promoters minus the percentage of Detractors, so it ranges from −100 to 100. Because it describes the overall relationship rather than a single moment, it works best as a periodic measure, for example once a quarter, tracked over time and compared across customer groups.
Customer Satisfaction Score (CSAT)
CSAT asks how satisfied the customer was with a specific interaction, such as a delivery, an onboarding call or a support conversation. It is most often asked on a five-point scale and reported as the share of satisfied responses (4 or 5) out of all responses, although companies use other scales too. CSAT is best sent soon after the interaction it asks about, while the customer still remembers it.
Customer Effort Score (CES)
CES asks how easy it was to get something done, such as resolving an issue or completing a return. It was introduced by the Corporate Executive Board, now part of Gartner, in the 2010 Harvard Business Review article "Stop Trying to Delight Your Customers". CES scales vary in length and direction, so state in every report which end of the scale means "easy". CES is most useful for processes, because effort is something a team can design out.
A new team does not need all three at once. Start with the metric that matches the decision you need to support: NPS for the health of the relationship, CSAT for specific touchpoints, CES for processes that customers complain are hard. Whichever you choose, the score tells you where to look, and the answer to the "why" question tells you what you will find there.
Where Customers Already Talk to You
Before adding new surveys, list the places where customers already give feedback. Each is a listening channel with its own strengths and its own bias, and knowing the bias is what lets you read the channels together.
Surveys are solicited feedback: customers answer the questions you chose, at the moments you chose. If your team runs surveys in a form builder, our guides to analyzing Typeform, Jotform and Hotjar data cover the export and analysis steps for each.
Support tickets, chat logs and call transcripts are the most detailed record of what goes wrong. Customers describe their problem step by step, and the conversation shows how the company responded. Almost everyone in this data had a problem, so it says little about overall satisfaction and a great deal about the causes of friction.
Reviews on app stores, marketplaces, travel sites and review platforms, along with conversations on social media and forums, are written without a prompt. They lean toward strong experiences, both very good and very bad, and they often mention things a survey never asks about, such as what the customer compared you with. They are also public, which means the same channel exists for your competitors.
Whatever the channel, keep the context with the feedback. A comment is far more useful when it comes with a date, the channel it came through, the product or plan involved and the customer's country or region. Those fields are what later make it possible to compare groups and periods. If you are choosing software to collect or manage these channels, our overviews of customer experience management tools, tools for tracking customer experience and tools for CX researchers compare the main options.
Segmenting Customers
An average hides most of what matters. Segmentation divides customers into groups that share a characteristic, so the team can see where the experience differs.
Demographic segmentation groups customers by attributes such as age group, income band or company size in a B2B business. Behavioral segmentation groups them by what they do: first-time buyers and repeat buyers, monthly and annual subscribers, or heavy and occasional users. Psychographic segmentation groups customers by values and priorities, such as those who choose mainly on price and those who choose mainly on quality; this information usually comes from a survey question rather than from your systems. Geographic segmentation groups them by country, region or city, which often reveals operational differences such as a slower carrier in one area.
Two practical points matter more than the choice of segmentation type. First, a segment can only be analyzed if it is recorded next to the feedback, as a column in the survey export or a field in the support ticket. Second, small groups mislead. If a segment has a dozen responses, one or two unusual answers can make it look very different from the rest, so treat findings from small groups as questions to explore rather than conclusions.
Turning Feedback Into Findings
When feedback arrives in dozens of comments a month, a person can read all of it and keep track of the patterns. Once it arrives in hundreds or thousands, reading stops working. Different people code the same comment differently, the same person codes it differently from one month to the next, comments that touch several topics get filed under one, and nobody can tell whether a difference between two regions is real or a coincidence of small numbers.
A sound routine solves most of this. Analyze on a fixed schedule, such as monthly for support data and after each wave for surveys. Use the same question and the same columns each time so periods can be compared. Let the themes come from what customers actually write rather than from a list of categories decided in advance. Give each theme its own sentiment, because one comment can praise the product and criticize the delivery. Compare segments and periods, and test whether a difference is large enough to be real before acting on it. Prioritize by combining how often a theme comes up, how negative it is and how much it differs between groups. Each of these steps is explained in detail in the analysis guide linked in the introduction.
Where Kimola Fits
Kimola is an AI platform that turns customer feedback into enterprise-grade market research. It analyzes reviews, survey responses, support conversations and social discussions the way a research team would.
A CX team starts with the data it already has. A survey export, a support ticket export or a file of call transcripts can be uploaded as a dataset in .xls, .xlsx, .csv or .tsv format, up to 30 MB. You choose the column that holds the feedback text and, optionally, a Date column and a URL column. Every other column in the file, such as plan, product, channel, country or the NPS score, is kept and becomes a filter and comparison group in the Report. Kimola can also read data from Zendesk, Intercom and Google Sheets through integrations, which are a paid-package feature. Feedback written in 30+ languages can be analyzed together in one Report.
Themes and sentiment from the data. Kimola finds the themes in the feedback without a predefined list of categories and names each one after what customers say. With multi-label classification, a comment gets every theme it mentions, and each theme gets its own sentiment.
Research questions asked of every comment. Dimensions ask the same question of every piece of feedback. For a CX team, Pain Points (what goes wrong or creates difficulty), Experience Stages (when in the experience it happens) and Unmet Needs (what customers want added or changed) are a natural starting set. Dimensions can be added when a Report is created or later.
Differences that are tested. Kimola crosses every theme and Dimension with the columns of the dataset and with time, and tests each combination. Only statistically significant findings become Signals: Differences show a topic that one group mentions more often or more negatively, and Trends show a topic whose mentions changed over time. If nothing in the data stands out, the Report has no Signals, which means nothing passed the test rather than that the analysis failed.
Recommendations checked against evidence. From the Actions screen, Kimola writes an Action Plan for a Profile, such as your company or one of its brands. It forms hypotheses from the Report, checks them against the feedback and leaves out recommendations the data does not support.
For a CX team that receives feedback continuously, the Enterprise package adds Spaces. A Space is a feedback channel that keeps growing, such as a call center or an NPS survey: the team adds a new file whenever new feedback arrives, every file is classified the same way so results stay comparable over time, and Signals run again after each file. A Space also shows an NPS calculated from what customers write, as the share of positive feedback minus the share of negative feedback. It is a different measure from the survey NPS described above, which comes from 0 to 10 ratings, and it does not replace it.
Kimola does not design or send surveys, host forms, run interviews or draw journey maps; it works after the feedback has been collected, alongside the tools that collect it. It is not a customer experience management platform and not a real-time social media monitoring dashboard, and it does not predict future behavior. If your team receives a few dozen comments a month, reading them directly may be all you need. Kimola has a free Starter plan with no credit card required, and paid plans start at $49 per month (pricing).
Learning From Competitors' Customers
Your own feedback tells you how your customers experience your company. It cannot tell you what customers value in the alternatives they considered, why some of them chose a competitor or what frustrates people across the whole market. Public reviews of competitors can. The customers of a competing app, hotel or product describe what they like, what goes wrong and often what they compared it with, and they do so in the same channels your customers use.
Treat this as research with a known bias. Reviewers are not a representative sample of anyone's customers; they lean toward strong experiences. The useful comparison is therefore not "their reviews are more positive than ours" but "their customers praise fast onboarding and ours complain about it", or "complaints about billing appear in every competitor's reviews, so this is a market problem rather than ours alone".
In Kimola, a Report can be created from links. Paste the URL of a product, business or app page on a supported platform, such as Amazon, Google Maps or the App Store, choose how many records to include and create the Report. To compare several competitors in one Report, the Add Multiple option accepts one URL per line, and the competitor each review came from becomes a column that Signals can compare across. For platforms without link support, Kimola's browser extension, the Airset Generator, collects feedback from 30+ supported sources into a dataset. When you write an Action Plan from such a Report, you can choose whether it is written from the perspective of your own offering, a rival or the market as a whole.
Sharing Findings With Other Teams
A finding only changes something when it reaches the team that owns the issue, whether that is product, marketing or operations, in a form that team trusts.
Every finding should travel with its evidence: how many comments mention the issue, which segment or period it concentrates in, how it has changed and two or three representative quotes. A sentence such as "complaints about missed delivery slots doubled in the northern region after the carrier change, in 140 comments over six weeks" invites action in a way that "customers are unhappy with delivery" never does. Give each finding an owner and agree when the CX team will check back.
Kimola Reports can be exported to PowerPoint, PDF and Excel for this purpose, and each action in an Action Plan shows the findings, Signals and feedback behind it, so the team receiving it can see where it came from. In the Enterprise package, Tasks turn a finding into work for a teammate with the evidence attached, and can be sent to Jira.
Closing the Feedback Loop
Closing the loop has two halves. The internal half is checking whether a change worked: after the product team fixes the onboarding flow, the CX team looks at the next period's feedback to see whether complaints about onboarding fell. Running the same analysis on the new data, with the same columns, makes this comparison direct, and in Kimola a Trend shows whether mentions of a topic changed between periods. In the Enterprise package, Kimola re-measures the feedback on a topic two and four weeks after the related Task is closed.
The external half is telling customers. People who took the time to explain a problem are more likely to keep giving feedback if they hear that something changed because of it. This can be a reply to the individual customer, a note in release notes or a message to everyone who raised the issue. Customers who see that someone listens keep talking.
Frequently Asked Questions
What should a new CX team set up first?
Start by listing the channels where customers already give feedback and agreeing who owns each. Then choose one core metric that supports a real decision, add a "why" question to it and make sure the feedback is stored with its date and the segment fields you will want to compare. A regular analysis routine and a way to share findings come next. New surveys and new tools can wait until these basics work.
Should a CX team start with NPS, CSAT or CES?
It depends on the question you need to answer. NPS describes the overall relationship and suits periodic tracking. CSAT measures satisfaction with a specific interaction, such as a delivery or support call. CES measures how easy a process was and suits teams working on friction in specific tasks. Many teams start with one, usually NPS or CSAT, and add the others once the first is running reliably.
Is customer experience the same as customer service?
No. Customer service is the help a company gives customers when they contact it, usually one interaction at a time. Customer experience is the customer's overall perception across the whole relationship, including marketing, buying, using the product, billing and support. Customer service is one important part of the experience, and support conversations are one of the richest sources of CX feedback.
Conclusion
A CX practice rests on a few foundations: clear responsibilities, one or two well-defined metrics with a "why" question, known listening channels with their context kept, segments recorded next to the feedback, a regular analysis routine, findings shared with their evidence and a loop that shows customers they were heard.
Most organizations already have the raw material in survey exports, support tickets and public reviews. A team that has that feedback can upload it to Kimola, keep the columns it already records and turn it into enterprise-grade market research, without designing a new study. You can create a free account to try it on your own data.



