CSAT — customer satisfaction — is the simplest survey in the book and the easiest to get wrong in Arabic. The metric is one question: “How satisfied were you?” on a short scale. Everything else is follow-up. The mistakes are rarely about the metric; they are about scale direction, vague wording, and asking ten questions when the answer to one would do. Here is a CSAT survey you can publish today, plus the reasoning for each choice.
The core question
Keep the scale to five points. A five-point scale reads cleanly on a phone, translates without ambiguity into Arabic, and gives you a stable percentage (top-two-box) to track over time. Anchor both ends in words, not just numbers, so “3” is not left to interpretation.
The template — six fields
| # | Question | Type | Required |
|---|---|---|---|
| 1 | Overall, how satisfied were you? | 5-point scale | Yes |
| 2 | What did we do well? | Long text | No |
| 3 | What could we improve? | Long text | No |
| 4 | Which part of your experience does this relate to? | Single choice | No |
| 5 | How likely are you to recommend us? | 0–10 scale | No |
| 6 | May we contact you about your feedback? | Email (conditional) | No |
Field 6 should only appear when the respondent opts in — a single consent checkbox that reveals the email field. Collecting contact details by default, without a reason, is exactly the pattern PDPL is written against.
Reading the result
Report CSAT as the share of 4s and 5s (top-two-box), not the raw average — an average of 3.8 hides whether you have a calm middle or a split room. The illustrative distribution below is the shape you want to watch: a healthy right lean, with the 1–2 tail small enough to read every comment behind it.
Keep it to one screen
The whole survey should fit on one phone screen without scrolling past the first two questions. CSAT response rates fall sharply once a respondent has to scroll to find the submit button. If you need more depth, branch into it from question 1 rather than stacking every question on everyone.