Customer Satisfaction Measurement Methods: 7 Approaches

Compare customer satisfaction measurement methods, adapt a survey template, and learn how to turn responses into measurable improvements.

Customer satisfaction measurement methods include surveys, interviews, review and complaint analysis, and the evaluation of customer behaviour data. Surveys use CSAT, NPS and CES metrics. The choice depends on the question: whether you need to assess a specific service interaction, willingness to recommend the company, or the difficulty of receiving the service.

Liels skaitlis 7 maigā rombā un teksts par veidiem, kā saprast, ko uzlabot klientu pieredzē.

Customer satisfaction measurement methods and selection criteria

Measurement is a systematic exploration of customer experience, aimed at understanding what to improve in the product, service or customer support. The goal is not simply to obtain a percentage, but to identify where the company's assumptions differ from customer needs.

Without this information, unclear communication or a complicated service step may go unnoticed. Use three criteria for selection: what decision needs to be made, which customers have the relevant experience, and how you will use the result.

Seven methods for measuring customer satisfaction

The following compares seven practical approaches, rather than a standard-defined list of methods. CSAT, NPS and CES are not separate data collection channels: they are metrics obtained through surveys.

Customer survey questionnaire: template, questions and scales

An editable template is available in the “Survey templates” section of VisiDati.lv, while Bibliotēka.lv provides adaptable CSAT and NPS questionnaires. Adopt the structure, not all the content: ask only about experiences known to the customer and information needed for analysis [biblioteka.lv, varam.gov.lv].

  • Which service did you use during your last interaction?
  • How satisfied are you with this experience overall?
  • How satisfied are you with the speed of receiving a response?
  • How satisfied are you with how understandable the information you received was?
  • How satisfied are you with the resolution of the problem?
  • What should be improved in this service?

For satisfaction, you can use a five-point scale from “very dissatisfied” to “very satisfied”. Name each category and, where necessary, add a separate “don't know/not applicable” option without including it in the satisfaction scale [biblioteka.lv].

Phrase questions neutrally and about one thing at a time: not “How fast and understandable was the response?”, but ask separately about speed and clarity. Avoid leading wording and double negatives; check comprehensibility before sending.

The questionnaire should be as short as the objective allows. For assessing a specific experience, keep only the necessary questions, while in-depth research may justify clearly reasoned follow-up questions.

CSAT: satisfaction with a specific service experience

CSAT is suitable after delivery or support. Using a 1-5 scale, the share of satisfied responses can be calculated as follows: number of ratings 4 and 5 / number of all valid ratings × 100. A comment helps explain the reason behind the result [biblioteka.lv].

NPS: customer willingness to recommend the company

NPS asks about willingness to recommend the company on a 0-10 scale. The score is the percentage of ratings 9-10 minus the percentage of ratings 0-6; ratings 7-8 are passives and remain in the total respondent count. It does not answer how successful a specific service interaction was [netpromotersystem.com].

CES: how easy it is for customers to receive a service

CES measures the effort invested by the customer, for example, in resolving a problem. It is useful when assessing service complexity; before aggregating results, record the scale direction and calculation so that a higher number is not mistakenly interpreted as a better result.

Customer interviews: exploring the reasons behind ratings

Choose interviews when you know where ratings are low but the reason is unclear. Ask customers to describe a specific case and ask clarifying questions; an interview does not replace a survey of a broader customer group.

Review and complaint analysis without a survey

Group complaints, comments, thanks and suggestions by topic. The absence of complaints is not proof of satisfaction: this channel provides only part of the overall experience. If reviews are collected in a CRM, formulate specific questions for your CRM about recurring topics.

In larger review projects, clients request centralised public review collection and analysis of sentiment, product category differences and possible customer churn signals.

Evaluating customer behaviour and service data

Evaluate service data alongside customer ratings. Response and resolution times are indirect signals, not the customer's satisfaction rating itself; use differences as a basis for further investigation.

How to choose a method for your company

Start with one question to solve, rather than implementing every metric. Choose CSAT to improve a specific experience, NPS to measure willingness to recommend, and CES to assess complexity; explore unclear reasons further through interviews.

Which customers to survey and how many responses are needed

Invite customers who have the experience being assessed. With a small customer base, you can contact everyone; with a larger one, plan sampling and representation of key groups [varam.gov.lv].

There is no universal number of responses required. Plan it based on the population, required precision, confidence level and groups to be compared; the calculation can be checked using the European Commission's JRC sample size tool. Required responses are not the same as invitations sent, and a large number of responses will not correct poor sampling.

When to send a survey and how to encourage responses

Send the CSAT question soon after the experience, but establish a regular cycle for overall attitude measurements. State the purpose, duration, deadline, use of responses and privacy conditions in the invitation; plan a reminder for non-respondents [biblioteka.lv].

In Latvia, Google search interest fell by 53% in the last 12 months through September 2026 compared with the previous 12 months; the peak was in November 2022. Interest is above average in January, March, May, October and November, but choose survey timing based on customer experience, not search seasonality.

Interese Latvijā sarukusi par 53%
Google search interest in Latvia by month from November 2022 to September 2026. Month with the highest interest = 100.

Which survey tool to choose

For a simple survey, consider VisiDati.lv or EUSurvey. Compare response limits, question types, languages, invitation management, data export and processing conditions; the tool itself does not determine whether the results will be usable.

Ready-made components offer NPS, CSAT and CES calculations, automated surveys and integrations, while larger survey projects also provide template management and no-code editors.

What real projects and public procurement include alongside a survey

Project requirements cover not only the visible questionnaire, but also administration, data processing and communication.

  • Customisable questionnaires: editable text, brand styling and ratings using faces, stars or numbers.
  • Management reports: NPS dashboards, rating summaries and grouping by customer or employee.
  • Connection with service delivery: automatic sending after a service and ratings alongside response and resolution times.
  • Access management: company accounts, user roles and links to reviews for a specific service.
  • Personalised communication: automatic email summaries, sometimes also PDF reports.
  • Data reliability: bulk import checks and protection against manipulated ratings.
  • Transferring paper data: extracting ratings, choices, text and handwritten entries from scans into spreadsheets.

In European public procurement, satisfaction and NPS software is purchased together with survey planning and routing, comment analysis, organisation-wide reports and personalised responses.

In European public procurement, postal service buyers purchase satisfaction and NPS management, while public telecommunications development buyers also purchase multichannel customer service with advanced artificial intelligence; these are different scopes of work.

How to aggregate results and interpret metrics

  • Keep the original data and check for errors, duplicates and missing responses.
  • Summarise rating distributions and percentages, and group comments by topic.
  • Compare customer groups and state the number of responses, response rate and limitations [methodology.eca.europa.eu].

There is no single “good” threshold for all companies. Compare it with your previous result and target, keeping the service, scale and customer group comparable; a small difference may be measurement uncertainty [biblioteka.lv, methodology.eca.europa.eu].

How to set improvement priorities and assess the impact of changes

Link low ratings to recurring comment themes and investigate the causes of problems. Study how aspects relate to the overall rating rather than automatically prioritising the lowest average [methodology.eca.europa.eu].

After changes, repeat the same questions and scales with relevant customer groups. Assess trends and comments together: an increase alone does not prove the impact of a change [biblioteka.lv, methodology.eca.europa.eu].

Frequently asked questions about costs, time and requirements

How much does a customer survey cost: free options and public tool prices

The VisiDati.lv free plan includes five surveys with up to 300 respondents each. The Professional plan costs EUR 53.50 for 12 months; the number of surveys and respondents is unlimited, but the respondent database is a paid service in both plans [visidati.lv].

Ekrānuzņēmums no ec.europa.eu
Screenshot from ec.europa.eu

The tool price is not the full cost of the research. Even a free tool requires time to develop questions, test them, collect responses and analyse the results.

How long it takes from survey preparation to conclusions

VARAM's simple survey method provides the following guideline, not a universal timeframe:

  • Question development: 3-4 hours.
  • Conducting the survey: 3-10 days or longer.
  • Analysing results: 4-8 hours [varam.gov.lv].

For a broader project, plan time according to the scope of work: in addition to the questionnaire, you may need to allow for interviews, integrations and report preparation. For integration requirements, also assess data exchange with the customer management system.

When customer satisfaction measurement is mandatory

The verified sources do not require all companies to conduct surveys. The obligation must be assessed based on the sector and applicable requirements: Paragraph 23 of Cabinet Regulation No. 399 requires public administration service owners to record recipient satisfaction metrics, including within delegated public administration tasks [likumi.lv].

Sources

LNB: From “feelings” to evidence (2026)
VARAM: Survey (2023)
Bain: Measuring Your Net Promoter Score
ECA: Analysis of a survey (2024)
VisiDati.lv: Prices
Likumi.lv: Cabinet Regulation No. 399 (2017)

Product and design team

UX designers and frontend engineers dedicated to crafting intuitive, high-performing interfaces that align product goals with user needs across B2B solutions.