A survey is one of the easiest things to send and one of the easiest to get wrong. Most surveys do not fail because people refuse to answer — they fail because the person who sent it never decided what they actually wanted to learn. So before you write a single question, answer four things: why, who, what and how. Get those right and the questions almost design themselves.
Why — start with the decision, not the survey
Every survey worth running exists to help you make a decision or catch a problem early. "General feedback" is a trap: it produces vague answers you cannot act on. So write down, in one sentence, what you would do differently depending on the answer. If you cannot, you are not ready to ask yet.
Purpose also sets the size. A quick post-visit pulse to catch a service dip is a different survey from a once-a-year deep dive into why customers stay. Knowing your why is what stops you asking twenty questions when three would do.
Who — the right people, while it is fresh
Feedback is only as honest as who you ask and when you ask them. Two rules carry most of it:
- Ask the people who actually had the experience, while it is still fresh — the diner leaving your café, the guest at the end of an event, the patient after the appointment. Memory fades fast; a survey sent a week later measures what people half-remember, not what happened.
- Do not only hear the extremes. The loudest five stars and the angriest one star will reach you anyway. The people worth hearing are the quiet middle who would never leave a review but will tell you the truth if you make it easy. Cast wide, keep it anonymous, and they will.
You do not need everyone. A steady trickle of honest responses beats a one-off blast. But be honest about how many you have: under about 30 responses, read the counts, not the percentages — "11 of 18 said" is truthful, while a percentage off a tiny base pretends to a precision it does not have.
What — ask less than you think, and ask it straight
The single biggest mistake is asking too much. Every extra question costs you completions, and a survey nobody finishes tells you nothing.
Anchor on two questions every time: an overall rating (how was it, 1 to 5) and the 0 to 10 "would you recommend us" — that second one is your NPS. Those two carry most of the signal. Then add only the handful that map back to your why.
Keep the wording clean:
- One idea per question. "Was the food fresh and well priced?" hides two answers in one — split it.
- Do not lead. "How much did you enjoy our excellent service?" tells people the answer you want. Ask "How was our service?" and let them tell you.
- Leave room for their words. One open question — "Anything we could do better?" — is where the gold is. It is where you learn the thing you did not know to ask about.
How — make it effortless to answer honestly
You have decided why, who and what. How you collect decides whether anyone actually responds.
- Meet people where they are. A QR code on the table or receipt, or a link by email, WhatsApp or SMS — whatever fits the moment. Phone-first, always; nobody fills in a form on a laptop after dinner.
- Match the length to the channel. As a rule of thumb: 3 to 5 questions for a QR at the counter (about 30 seconds), 8 to 12 for an emailed survey where people expect to spend a minute or two, and 6 to 10 for a post-event form while it is fresh. Past roughly 20 questions, completion falls off a cliff.
- Anonymous by default. People are far more honest when they are not signing their name to a complaint. Only ask who they are if you genuinely need to follow up — and say why.
- Close the loop. Collecting is half the job. The point is to read what came back, fix the one or two things that most move how people feel, and, where you can, let people see that you did. Feedback that visibly changes something is the best reason to give it again.
The short version
Why: know the decision before you ask. Who: the right people, while it is fresh — not just the loud ones. What: two anchors, a few honest questions, and one open box. How: phone-first, short, anonymous, and always followed by action.
That is the whole discipline — and it is exactly what Vitalz is built to make easy. It drafts a right-sized questionnaire for your trade, collects responses by link or QR, and turns them into an honest report that tells you what to fix first — with no invented numbers and no busywork, just a clear read on what your customers actually think.