Satisfaction survey
A short questionnaire sent after a contact to turn the customer's perception of that episode into a comparable score.
Also known as: customer satisfaction survey · satisfaction questionnaire · support satisfaction survey · satisfaction survey form
Try CRPRO free for 7 daysA satisfaction survey is the questionnaire applied right after a contact — support, purchase, delivery, a technical visit — to turn perception into a score. It does not measure what the business delivered; it measures what the customer thought they received. The two often diverge, and it is exactly that divergence that matters.
The three formats that cover almost every case
| Format | Question | Scale | What it measures |
|---|---|---|---|
| CSAT | "How do you rate this support interaction?" | 1 to 5 | one specific interaction |
| NPS | "From 0 to 10, how likely are you to recommend the company to a friend?" | 0 to 10 | the relationship with the brand |
| CES | "Was it easy to resolve what you needed?" | 1 to 5 | the effort the customer had to put in |
Running all three at once is the most common mistake beginners make. Pick one: if the goal is fixing support, CSAT; if it is tracking the brand over the year, NPS.
When to send it
The right trigger is the end of the episode, not a date on the calendar. CSAT goes out minutes after the conversation closes, while the customer still remembers the agent's name. NPS runs on a cycle — quarterly or twice a year — and never the day after a problem, or you measure the problem and think you measured the relationship. A survey that arrives three days later gets a made-up score.
How many questions fit
One closed and, at most, one open. The closed one gives you the number you track; the open one ("what was missing?") gives you the reason, which is the actionable part. Every extra question tanks the response rate, and a ten-item form usually only gets answered by whoever is very angry or very grateful — exactly the sample you do not want.
Where a satisfaction survey fails
- Response bias: whoever answers is whoever holds an extreme opinion. The silent majority is left out and the average does not represent the base.
- A deceiving average: 60% giving a 5 with 40% giving a 1 produces 3.4 — the same average as a company where everyone gave a 3, and those are opposite realities. Look at the distribution, not just the average.
- A survey with no follow-through: if nobody calls whoever gave a 1, the survey turns into a ritual and the customer learns to ignore it.
- Asking before resolving: with the ticket still open, you are measuring anxiety.
- An agent asking for "give me a 5": once the score becomes an individual target, it stops measuring the customer and starts measuring the team's skill at asking for scores.
Satisfaction survey on WhatsApp
WhatsApp gets far more responses than email because the person does not need to leave the app or open a link. Three practical precautions: once 24 hours have passed since the customer's last message, sending it requires a Meta-approved template; the question needs to fit on one screen, with options as buttons or a numbered list; and the score needs to go back into the contact's record, or you accumulate statistics with no owner and cannot call back whoever complained.
Frequently asked questions
How many questions should a satisfaction survey have?
One closed and, at most, one open. The closed one generates the number you track over time, and the open one explains the reason for the score. Every additional question tanks the response rate, and long forms end up answered only by whoever is very upset or very satisfied.
What is the best time to send the survey?
Right after the support interaction ends, still the same day, while the person remembers the details. If the survey arrives days later, the customer answers based on their overall impression of the brand and not the episode you wanted to measure. The exception is NPS, which measures the relationship and so runs on a quarterly or twice-yearly cycle.
Can I send a satisfaction survey over WhatsApp?
You can, and the response rate is usually much higher than email's because the person answers without leaving the app. If more than 24 hours have passed since the customer's last message, sending it requires a Meta-approved template. Also keep a record of anyone who asked not to receive that kind of contact anymore.
How do I calculate the survey's result?
It depends on the format: with CSAT, divide positive responses by total responses; with NPS, subtract the percentage of detractors from the percentage of promoters. In either case, look at the distribution of scores, not just the average, because equal averages can hide opposite realities. Log how many people responded, since small samples swing a lot from one week to the next.
Does a satisfaction survey need customer authorization?
The same logic as other contacts applies: you need a legal basis to process the data and a clear way for the person to decline future sends. Since the survey stems from an existing consumer relationship, it is usually treated as relationship communication rather than marketing. The practical care is respecting anyone who asks to stop and not using the survey as an excuse for an offer.
Related pages
Related glossary terms
CSAT
An indicator that expresses, as a percentage or an average, the score a customer gives to a specific interaction right after it ends.
NPS (Net Promoter Score)
A loyalty metric calculated from the 0-to-10 question about the likelihood of recommending, subtracting the percentage of detractors from promoters.
Likert scale
A response scale with ordered points, from strongly disagree to strongly agree, that converts opinion into a comparable number.
Customer satisfaction
The result of comparing the expectation a customer had against the experience they perceive they received.
Support SLA
An agreed deadline for replying to or resolving a support request, used as an operational target and measured against each conversation's real time.