Survey questions that get answered honestly

A guest receives an email the morning after dinner: "We'd love your feedback!" It opens a survey with twenty-four questions. How would you rate the cleanliness of the toilets? How likely are you to recommend our wine list to a colleague? How satisfied were you with the temperature of your main course? By question seven, the guest closes the tab. The few who finish are either delighted or furious, and the venue learns little about the guests in between.

Surveys can be one of the most useful sources of guest feedback, or one of the most misleading. The difference lies in the questions: how many, which ones and how they are worded. A short survey with well chosen questions gets honest answers from many guests; a long one with leading questions gets flattering answers from a few.

This guide covers six survey questions worth asking after a visit, the two kinds of question that bias every answer, the one open box that earns its place, and how to keep the whole thing short enough that guests finish it. Every venue and figure in it is an example.

Most surveys fail because they are too long, too vague or too eager to please. A guest asked twenty questions answers the first few carefully and the rest at random, or not at all. A guest asked questions that hint at the right answer gives the right answer.

The result is a set of feedback scores that look good and mean little. Average ratings drift towards the top of the scale, the comments come only from the happiest and the angriest guests, and the venue cannot tell what to change.

A survey that works is short, neutral and focused on what the venue can act on.

The guests who answer surveys are not all your guests. Very happy and very unhappy guests answer more often than the ones in between, so the answers lean towards the extremes. Regulars answer more often than first-time guests.

Keep that in mind when reading the results. A survey's average says something about the guests who answered, not necessarily about every guest. The more guests answer, the closer it gets to the truth, which is another reason to keep the survey short.

The single most important rule is to keep the survey short. Three to six questions, answerable in under a minute, is enough for most venues. Every question added lowers the number of guests who finish.

Ask only what you will act on. If you will not change the wine list whatever guests say, do not ask about it. If you want to know about the food, the service and whether guests will return, ask about those and nothing else.

A guest who finishes a short survey and sees it was quick is more likely to answer the next one too.

Start with one overall question: how would you rate your visit overall? Use a simple scale, such as five points from very poor to excellent, with words rather than only numbers, so every guest reads the scale the same way.

The overall question gives you the headline figure for each visit and each period, and it lets you compare the other questions against it. A guest who rates the visit poor overall but the food excellent has told you the problem was elsewhere.

Next, ask about the three things guests judge most: the food, the service and the atmosphere. Use the same scale as the overall question, so the answers can be compared.

These three questions show where the overall rating comes from. A venue whose service is rated well below its food knows where to look. One whose atmosphere is rated poorly may need to look at noise, lighting or how busy the room was.

Keep each question about one thing only. "How was the food?" is clearer than "How was the quality and value of the food?", which asks two things at once.

If value matters to your venue, ask about it separately: how would you rate the value for money of your visit? Guests often feel differently about value than about quality. A guest can love the food and still feel it was expensive.

Value is worth asking about when you are considering prices, portion sizes or promotions. If it is not something you will act on, leave it out and keep the survey shorter.

The last scored question asks about the future: how likely are you to visit again, or to recommend us to a friend? The second form is the basis of the net promoter score, which sorts guests into promoters, passives and detractors by their answer on a scale from nought to ten.

Net promoter score can be useful as a single figure to track over time, but it has limits in a restaurant: many guests rarely recommend restaurants to anyone, and the score can swing with small numbers of responses. Use it as one signal among several, not as the only measure.

After the scored questions, offer one open box: is there anything you would like to tell us? Or, more pointedly: is there one thing we could have done better? An open box catches what the scored questions miss: a cold plate, a rude remark, a server who made the evening special.

One open box is enough. Several open boxes, each asking for detail on a different topic, turn a quick survey into an essay most guests will not write.

Read every answer. The open box is where the most useful feedback usually hides.

Ask about the guest's visit, not about the venue in general. "How was the food on your visit?" invites a specific answer; "How do you rate our food?" invites a general impression that may be weeks old.

Linking the survey to a specific visit, with its date, also helps you act on the answers. A complaint about slow service on a particular Friday can be checked against what happened that night.

The order of questions affects the answers. Asking about specific things first, then the overall experience, can pull the overall rating towards the specific ones. Asking the overall question first gives a cleaner headline figure.

Start with the overall question, then the specific ones, then the open box. Keep the order the same every time, so this month's answers can be compared with last month's.

The first kind of question that biases every answer is the leading question: one that suggests the answer it hopes for. "How much did you enjoy your meal?" assumes the guest enjoyed it. "How delicious was our new menu?" assumes it was delicious.

Leading questions produce flattering answers, because most guests are polite and go along with the wording. The scores look good and mean little.

Use neutral wording instead: "How would you rate the food?" with a balanced scale from very poor to excellent.

The second kind is the double question: one that asks about two things at once. "Was the food hot and well presented?" "Were the staff friendly and efficient?" A guest whose food was hot but badly presented cannot answer honestly.

Double questions produce muddled answers that cannot be acted on. If a guest rates "friendly and efficient" as average, you do not know which part was the problem.

Split them into separate questions, or choose the one that matters more.

Use the same scale for every scored question, and label it in words: very poor, poor, okay, good, excellent. Words are read the same way by most guests; bare numbers are not, because some guests think five out of ten is fair and others think it is terrible.

Avoid scales with no middle, which force guests to lean one way, and scales with too many points, which make the answers harder to read. Five points is enough for most questions.

Offering a reward for completing a survey, such as a discount on the next visit, raises the number of responses but can change the answers. Guests who expect a reward may answer more kindly, or answer quickly without thinking.

If you offer anything, make it modest and the same for every guest, whatever they say. Never make a reward depend on a positive answer.

When the survey arrives affects the answers too. A survey sent within hours catches the guest's immediate impression; one sent a week later catches a memory, often softened or sharpened by time.

Send the survey soon after the visit, usually the next morning, and keep the timing the same for every guest, so the answers can be compared fairly.

Most surveys arrive in feedback emails after the visit. The email matters as much as the questions. Send it soon, while the visit is fresh, usually within a day. Keep the subject short and the email shorter. Thank the guest, say how long the survey takes and put the first question, or a clear button, right in the email.

Our guide to the post-visit email, sent the next morning covers the timing and the tone.

Some guests want a reply; others want to give feedback privately. Ask which, at the end of the survey: would you like us to follow up with you, and if so, how? Guests who want privacy should be able to give feedback without being identified.

That one question changes what you can do with the answers. A guest who asks for a reply after a poor visit is giving you a chance to put it right; one who asks for privacy is giving you honest feedback they might not otherwise share.

A handful of responses tells you little; a few dozen begin to show patterns. Track your feedback scores over weeks and months rather than reacting to a single response.

Look at the number of responses as well as the scores. A rise in the average from a smaller number of responses may simply mean fewer unhappy guests bothered to answer.

Surveys can be offered at the table, at the end of the meal, or by email after the visit. Each reaches different guests: the table catches guests in the moment, while the email reaches guests once they have had time to reflect.

Many venues use both. Our guide to feedback at the table, asked properly covers the in-person side.

A full survey belongs after the visit, but a single question can be asked at the table: how would you rate tonight overall? A one-question survey on the bill, a card or the guest's phone catches guests who would never open an email.

Keep it to one question and an optional comment. Anything longer belongs in the email.

Compare feedback scores like with like: the same days, the same services and similar numbers of responses. A score from a quiet Tuesday lunch is not comparable with one from a busy Saturday dinner.

Look at the trend over several weeks rather than week to week. Small changes from week to week are usually noise; a steady movement over a month or two is a signal.

A survey is only worth sending if the answers are read and acted on. Read every response, especially the open box. Each week or month, look at the scores by question and the themes in the comments, and choose one or two changes.

Tell guests and staff what changed because of feedback. Guests who see their feedback acted on are more likely to give it again. Our guide to acting on feedback: the Monday meeting item covers turning feedback into changes.

Some questions do more harm than good. Avoid asking guests to rate individual members of staff by name, which can feel unfair to staff and awkward to guests. Avoid asking for personal details the survey does not need. Avoid asking about things you will not change.

Avoid asking guests to leave a public review in the same breath as asking for private feedback. Mixing the two can make guests feel the private feedback is only a filter for the public one. Our guide to asking for a review without begging for one covers public reviews separately.

Most guests open a survey on their phone. Make sure it works well on a small screen: large buttons for the ratings, one question per screen or a short scrolling page, and an open box that is easy to type in.

A survey that is awkward on a phone loses guests at the first question. Test it on several phones before sending it to anyone.

Before sending a survey to guests, ask a few colleagues or friends to complete it. Time how long it takes, watch where they hesitate and ask whether any question confused them.

A question that confuses a colleague will confuse guests too. Fix it, test again and only then send it out.

If many of your guests speak another language, offer the survey in it. A guest who struggles with the language of a survey answers quickly or not at all.

Keep translations simple and check them with a native speaker. A badly translated question can change its meaning and bias the answers.

Ask only for the personal details the survey needs. If you do not need a guest's phone number, do not ask for it. Explain how you will use the answers and any contact details, and keep them safe.

Guests are more honest when they trust that their feedback will be used properly and their details will not be misused.

When a survey leads to a change, say so: a line in the next email, a reply to the guest who raised it, or a note at the foot of the survey itself. Guests who see that their answers led somewhere are more likely to answer next time, and to answer honestly.

An invented bistro sends a twenty-question survey after every booking. Few guests finish it, and the average rating sits near the top of the scale. The open boxes, scattered through the survey, are mostly empty.

The bistro cuts the survey to five questions: overall, food, service, atmosphere and how likely the guest is to return, each on a five-point scale with words, plus one open box asking what could have been better, and a question about follow-up. The email goes out the next morning with the first question in the email itself.

Responses rise sharply. The average rating falls a little, which the owner realises is more honest, and the open box reveals a recurring complaint about the wait for the bill, which the team fixes within a week.

In servearound, the feedback form is short and fixed: an overall rating from very poor to excellent, then ratings for the food, the service and the ambience, a box for comments and the guest's name and contact details if they want to give them. Guests then choose whether the venue may follow up, by email, by phone or either, or keep their feedback private, in which case they are never identified to you. The questions cannot be changed, so anything else you want to ask needs a survey of your own.

Guests reach the form from a link after their visit or from the menu they open with the table's QR code once their order is in, and each visit can be answered once. The team sees the responses, the average rating and how many guests are happy to be contacted, and can send a follow-up email to those who asked for one. See Guest CRM.

Three to six, answerable in under a minute.

Overall experience, food, service, atmosphere, value if it matters, and whether the guest would return or recommend.

One that suggests the answer it hopes for, such as "How much did you enjoy your meal?".

One that asks about two things at once, such as "Was the food hot and well presented?".

As one signal to track over time, but not as the only measure.

One, asking what could have been better. It is often where the most useful feedback hides.

Within a day of the visit, while it is fresh.

Read every one, look for patterns, choose one or two changes and tell guests and staff what changed.

Survey questions get honest answers when there are few of them, they are neutral and each asks about one thing. Ask about the overall experience, the food, the service, the atmosphere, value if it matters and whether guests would return, each on the same five-point scale with words, and add one open box asking what could have been better.

Avoid leading and double questions, send the survey within a day, ask how guests want to be followed up, track the scores over time rather than reacting to single answers, and make sure every response is read and acted on.