01 / UNDERSTAND
What is Customer Satisfaction Score?
Customer Satisfaction Score captures a person’s reported satisfaction with a named experience, product or service. One common approach asks for a rating on a five point scale and reports the percentage selecting the top two satisfaction categories. Declare the scale and calculation, because CSAT implementations vary.
Satisfaction is different from task success, perceived effort or recommendation intent. A person can complete a task, find it difficult and still be satisfied with the overall result.
Why use Customer Satisfaction (CSAT)?
A focused satisfaction question can reveal whether an interaction meets expectations, especially when an optional explanation is collected. It helps a team review a particular service moment without pretending to measure the whole relationship. The designer’s task is to connect the response to a specific experience and avoid interpreting a favourable score as proof that the underlying process is sound.
When to use it
USE IT WHEN
A defined interaction has ended and people can reasonably assess it.
The team wants a consistent feedback signal across comparable experiences.
There is capacity to investigate explanations and review missing responses.
When to choose another method
Use Customer Effort Score (CES) when the question concerns effort.
Use Task Success Rate for observed completion.
Use Net Promoter Score (NPS) for recommendation intent about a relationship.
What you’ll need
Name the experience and choose a point when its outcome is knowable. Agree response labels, invitation eligibility and handling of missing answers. Keep an optional reason question and record both the number invited and valid responses. Include people whose interaction failed where you can invite them appropriately; a survey available only after successful completion excludes an important experience.
What a CSAT measurement contains
Keep the elements that help answer your question and support a decision.
The precise interaction or outcome being rated.
When respondents can judge that experience.
Response labels and direction, held consistent.
Who can receive the invitation, including failed interactions.
The declared rule for satisfied responses and valid ratings.
Optional reasons and evidence needed for interpretation.
02 / PUT IT INTO PRACTICE
How to measure and interpret CSAT
Keep evidence and assumptions distinct. Revisit your work when new information changes your understanding.
01
Define the moment
Choose an experience people can identify without guessing which part you mean.
Ask about arranging a repair visit, not the whole housing relationship.
02
Choose timing
Allow the outcome to become clear before asking for an assessment.
An accepted request is not yet a completed repair.
03
Specify the scale
Use clear ordered satisfaction labels and record the scoring convention.
On a five point scale, define satisfied as ratings four and five.
04
Invite consistently
Avoid asking only the easiest cases or most favourable customers.
Provide a relevant route for failed booking attempts where possible.
05
Calculate transparently
Divide satisfied valid responses by all valid ratings under the declared rule.
Eighteen satisfied responses out of thirty equals 60% CSAT.
06
Read the distribution
Keep neutral and dissatisfied responses visible and review their explanations.
A rating of three can conceal a different issue from a rating of one.
07
Act on a question
Use feedback to select an investigation and compare the same experience after change.
Examine why a confirmed visit still feels uncertain.
Example: satisfaction with arranging a repair visit
Fictional scenario: 60 people are invited after a booking attempt and 30 provide valid ratings. Eighteen select satisfied or very satisfied on a five point scale. The declared CSAT calculation is 18 divided by 30, or 60%. The response rate is 50%. These are invented figures.
Comments distinguish a clear booking interaction from uncertainty about whether the contractor will attend. A designer should not assume the first screen is responsible for the whole rating. The example displays the experience boundary, response count and distribution before the headline percentage.
The next question concerns confirmation credibility. The team would inspect operational evidence and interview people about what they believed was guaranteed. A higher score after new copy would still need checking against whether visits actually happen as promised.
03 / BELOW THE SURFACE
Below the surface: satisfaction can rise when expectations fall
The same experience can receive a different rating when people expect less.
BOUNDARY
Which experience did the person actually rate?
Name the event and outcome.
EXPECTATION
Against what promise was it judged?
Read the reference point in comments.
ABSENCE
Who could not or did not respond?
Check unsuccessful experiences and invitation access.
A person who has repeatedly been disappointed may describe an ordinary appointment as satisfactory because it finally happened. Another person may judge the same delay harshly because the service promised immediate resolution. A satisfaction score records an appraisal, not an absolute measure of service quality.
When reviewing comments, note the expectation being used and whether it came from the service, previous experience or another provider. Do not improve a score by quietly lowering the promise. A more honest promise can be useful, but it should also support an acceptable outcome.
This editorial reading pairs satisfaction with relevant operational and task evidence. If ratings improve while delays grow, investigate who responded and what they now consider normal. The contradiction can be more informative than the average.
Customer Satisfaction (CSAT) companion worksheet
The original two page companion records the rated moment, invitation, scale, distribution and CSAT arithmetic. Its review page connects expectations and missing voices to a follow up. It supports a declared scoring convention rather than presenting one CSAT variant as universal.
Practical tips
State the experience in the question.
Show valid responses and invitation counts.
Review satisfaction together with whether the promised outcome occurred.
Common mistakes
Treating missing answers as neutral satisfaction.
Comparing surveys with different labels or timing.
Using a favourable score to dismiss an operational failure.
What you’ll get
A bounded account of satisfaction with a specific experience. Use the distribution and reasons to investigate one consequential gap, then review both reported satisfaction and the outcome people received.