A guest satisfaction survey has a rival that most other surveys do not: the public review. Whatever the booking platform is, it will invite the guest to rate the stay soon after checkout, and that rating is permanent, visible and weighted into how often the property gets shown. A survey that arrives after the review has been posted is collecting history.
That single fact reorganises everything about how the survey should be built. The question set matters, and it matters less than the timing and what happens in the twenty minutes after a low score lands. This guide covers the questions worth asking about a stay, where to place the survey so it is actually seen, and how to route a complaint to a named person while the guest is still on the property or still willing to talk.
Ask during the stay, not only after it
The most valuable feedback in hospitality is the feedback that arrives while the guest is still in the building, because that is the only window in which the problem can be solved rather than apologised for.
A guest whose room was not made up can be given a made up room. The same guest, surveyed four days after checkout, can only be given a discount voucher and an explanation, and will have already told the internet. The cost difference between those two outcomes is large and it is entirely a question of timing.
This argues for two surveys rather than one. A very short in stay check, one or two questions, delivered somewhere the guest will encounter it naturally. And a fuller post stay survey for the detail that only makes sense in retrospect, such as value for money and whether they would book again.
The in stay check has to be genuinely short. A QR code on the desk card that opens a single question, "Is everything as it should be?", with a yes, a no and a text box, will get answered. The same card pointing at a twelve question form will not. The purpose is not measurement. It is to give a guest who would otherwise say nothing a two second way to raise a hand.
For the post stay survey, send it the same day as checkout or the morning after. Later than that and the public review has usually already been written, which means the survey has lost its most useful function.
The question set for a stay
Group the questions by the part of the stay they cover, and keep the whole thing to eight rated items at most.
Booking and arrival. How easy was it to book, and how was check in? Arrival is where a stay is most often decided, and it is the part staff see least, because the guest who waited twenty minutes at a busy desk rarely mentions it at the time.
The room. Cleanliness, comfort of the bed, quietness, and whether everything worked. Keep these separate rather than folding them into one "room quality" rating. A property can be spotless and unbearably noisy, and a single combined score hides which of the two needs money spent on it.
Staff. Helpfulness and responsiveness, as two questions if the property has both a front desk and housekeeping. This is the area where scores are highest and most useful, because staff feedback that is specific and positive is worth passing on by name.
Value. Whether the stay felt worth the price paid. This correlates with the rate the guest actually paid, not with the quality of the room, so read it against the booking channel rather than in isolation.
Return and recommend. Would they stay again, and would they recommend the property. Two questions that look alike and are not. Business travellers often return without recommending, and guests on a one off trip often recommend without any intention of returning.
Then two open questions, and only two. What stood out, and what should be fixed. The temptation to add "any other comments" as a third box should be resisted, because it takes answers away from the two specific ones.
What not to ask
Leave out anything the property cannot change and anything the guest cannot judge. Questions about the neighbourhood, the weather or the local transport produce scores that move a property average around for reasons nobody can act on. So do questions about things the guest never saw, such as the booking system used behind the desk.
Two more to avoid. Do not ask for a rating of the property overall in addition to the individual items, because the overall figure then competes with the items and the two will disagree in ways that get argued about. And do not ask about price and value in the same question. Price is a fact the guest already knows, and value is a judgement about it, so combining them produces an answer that could mean either.
Where the survey lives
Placement decides response rate far more than question wording does.
| Channel | When | Response rate | Best for |
|---|---|---|---|
| QR code in room or on the desk card | During the stay | Low, but the responses are urgent | Catching problems while they can be fixed |
| Tablet or card at checkout | At departure | Moderate | Short stays where email is not collected |
| Email, same day as checkout | After | Moderate to good | The full question set |
| SMS, a few hours after checkout | After | Good, if the number was collected | One or two questions only |
| Link in the booking confirmation | Before | Very low | Not worth using for satisfaction |
A property running one channel well beats a property running four badly. The email sent on the day of checkout is the workhorse, and the in room QR code is the one that pays for itself the first time it saves a review.
One technical point that decides whether any of this works: the survey must open without an account and finish on a phone. A guest will not create a login to complain, and most of these links are opened on a phone while walking to a car. One question per screen suits that situation better than a long scrolling page.
Route the bad ones before anything else
A survey is worth building when a low score has a named destination. Without that, the property has bought a measuring device for a problem it is not going to act on.
Decide three things in advance. What counts as low, which means writing down the threshold rather than leaving it to judgement. Who receives it, by name and by role, including who covers nights and weekends, since a Saturday complaint that waits until Monday has already become a review. And what that person is authorised to do without asking, because a duty manager who has to seek approval for a room change will not make the call in time.
Then the mechanics. Each response needs an owner and a status, so that a complaint seen by two staff members is not either answered twice with different offers or left because each assumed the other had it. The reply has to be sent from the same place the response is read, so what was promised to the guest stays attached to the record instead of sitting in one person's sent items.
This is the difference between a form that emails a copy to a shared mailbox and a form tool with response management behind it. The shared mailbox works until two people are looking at it, which in hospitality is immediately, because the whole point is that the next shift picks up what the last one started. Looking at what the tools after the form actually do matters more here than comparing question types.
A second requirement follows from repeat guests. Responses keyed on the email address should join up, so a returning guest arrives with their previous stay and previous complaint already visible. Recognising that a guest is on their fourth stay changes how the reply should be written.
One list across every property and unit
Groups with more than one property, and short let operators with more than one unit, hit the same problem: the feedback splits by property and the picture never comes back together.
The pattern that holds up is a single form with the property or unit as a field, rather than one form per property. The guest sees one question more. The operator gets one list that can be filtered by unit, compared across units, and exported as a single file, instead of five forms whose results have to be reconciled by hand every month.
The exception is where each property has genuinely different questions, such as a hotel with a restaurant alongside a serviced apartment block with none. Even then, duplicating one form and editing the questions keeps the structure identical and the exports compatible, which matters when the monthly report is assembled.
Internal notes belong in the same record too. What the guest said, what was offered, whether it was accepted, and whether the underlying fault was fixed are four separate facts, and the last two are invisible to the guest. Keeping them as fields on the response rather than as a comment thread means they appear as columns in the export and can be counted. Several of the questions people ask cover exactly this, the difference between what the respondent sees and what the team records afterwards.
Read the results by unit and by cause
Two habits make the results usable.
Break the scores down by room, unit, shift and booking channel before looking at any average. Hospitality problems are rarely evenly distributed. Three rooms above a delivery bay will drag a noise score down, and no amount of staff training will move it. A property wide average of 4.1 out of 5 tells you nothing about which room to soundproof.
Then separate the fixable from the structural. A slow check in at peak arrival time is a staffing decision. A thin wall is a building. Both show up as low scores in the same column, and treating them the same way produces a maintenance list nobody can act on. Tagging each complaint by cause as it arrives, rather than in a monthly review, is what makes this possible, because the person reading the response knows which it is and the person compiling the report two weeks later does not.
Finally, count outcomes rather than responses. How many low scores were contacted, how fast, and how many resulted in a fix that can be pointed at. A rising satisfaction score with no list of changes behind it usually means the mix of guests changed.
What to change first
Put a one question check in the room with a QR code, and give every response below the threshold a named owner with authority to act before the guest checks out. That single change recovers reviews that would otherwise be written. Once it is running and the volume starts to hurt, see what one list from response to reply looks like in Halict.
Q1. When should a guest satisfaction survey be sent?
Send the full survey on the day of checkout or the following morning, because booking platforms invite a public review soon after departure and a survey that arrives later is collecting history. Alongside it, run a one or two question check during the stay, which is the only window in which a problem can still be fixed rather than apologised for.
Q2. How many questions should a hotel survey have?
No more than eight rated items and two open boxes for the post stay survey, and one or two questions for the in stay check. Cleanliness, noise, comfort, arrival, staff and value are the items that earn their place. Anything beyond that lowers completion without changing a decision.
Q3. Should the survey be anonymous?
Identify the guest and ask permission to follow up, otherwise a complaint cannot be answered. Anonymity produces slightly more candid comments and removes the ability to recover the relationship, which is the more valuable of the two in hospitality. If anonymity is promised for a particular survey, honour it completely.
Q4. How do you stop guests being asked for feedback too often?
Set one rule and enforce it in the sending tool rather than by convention: one post stay survey per stay, and no survey to a guest who already answered one within the same trip. Operators with several properties need the rule applied across all of them, since a guest moving between units in a group can otherwise receive three requests in a week.
Q5. Is one form enough for several properties or units?
One form with the property or unit as a field is easier to run than one form per property. The results stay in a single list that can be filtered and compared, and the monthly export comes out as one file. Duplicating a form is only worth it when the questions genuinely differ, such as a property with a restaurant compared with one without.