Retail Customer Experience Survey Template
Twelve questions about one visit to one store, answerable by somebody who bought nothing and has no account with you. Take it here first, then open it in the editor and put the link where people leave rather than where they pay.
- 12questions
- About 2 minto answer
- Question 1names the store
- Question 2buyers and non-buyers
The 12 retail customer survey questions
Every question, the scale it uses, and what a low answer is pointing at. Copy one block, copy all twelve, or open the whole thing in the editor.
Which store, and whether they bought
Two questions that decide how every answer below is read, and neither can be recovered afterwards. A rating does not say whether the person left empty handed, and it does not say which of your stores they were standing in.
- Which of our stores were you in?Rename the four options to your own stores before you send it. Read the answers inside each store: an average across all of them describes no store you own.
- Did you buy anything on this visit?The fork. Split every rating below on this first. The third option keeps a collection visit out of the non-buyer pile, where it would look like a lost sale and was not one.
The errand
A visit starts as an errand. These two separate a store that did not have the thing from a store that had it somewhere the shopper could not find it.
- The store had what I came in for.Low from a non-buyer is the plainest answer on the page: they came for something and you did not have it.
- The store was laid out so I could find my way around.Low here beside a high answer above means the stock was in the building and the shopper never reached it. That is a layout job, not a buying job.
The people and the place
Three judgements about what the store was like to be in, each worded against what the shopper expected rather than against a standard they have never seen.
- Someone who works there was free when I wanted help.Staffing as the customer met it, not as the rota describes it. Read it against the hour and the day the visit happened.
- The person who helped me knew the products.Optional, because plenty of visits involve nobody at all. Forcing it puts an invented score against whoever happened to be on the floor.
- The store was as clean as I expect.Against expectation, so a discount store and a jeweller are each judged on their own terms.
The till, and the question nobody else asks
One rating only a buyer can answer, and the open question this whole survey exists for.
- Paying was quick enough.Optional, because a third of the people answering never reached a register. Judged against expectation rather than in minutes: nobody can count minutes from memory.
- If you left without buying anything, what stopped you?The stem names its own condition, so a buyer skips it and a non-buyer cannot miss it. Every other channel a store can use would have lost this person before they answered.
Their own words, then the close
One open box for anything the ratings did not catch, then two ratings to end on.
- What is the one thing you would change about this store?One thing, not a list. Short answers are the ones a store manager reads on a Monday.
- I would go back to this store.Return intent, not the 0 to 10 recommend question. Coming back is the thing one store can act on.
- Taking it all together, this visit went the way I wanted.The close. A high answer sitting over a weak middle is not a contradiction: it says somebody let you off.
Swap-in items for what your store actually has
Add the one that applies and take out anything the shopper could not have seen. Four is the ceiling: past sixteen items this stops being answerable in a car park.
- Trying things on was straightforward.Where a browsing visit can end without a transaction, which is the part no till-side question reaches.
- The prices on the shelf matched what I paid.A pricing-integrity item. Put it beside question 8 and read the two together.
- I could try the product out before deciding.For anything a shopper expects to handle. A low answer here beside a low question 3 is a display problem, not a stock problem.
- Parking was easy enough.Part of the visit for a retail park or an outlet, and part of nobody else on this page.
Editing it. All twelve are already loaded. Rename the store options in question 1 to your own stores, drop the swap-ins you do not need, and send one link today from the quick survey maker. Keep questions 1 and 2 whatever else you change: without them the other ten are one average of several stores and two kinds of visit.
What is deliberately missing. No age band, no gender and no income question: none of the three changes what a store manager does on Monday. There is no scored satisfaction percentage either, because that arithmetic has its own page in the CSAT survey template. The 0 to 10 recommend question and its bands belong to the net promoter survey. And when what you are measuring is how hard one request was to resolve, the customer effort score survey carries the scored effort item.
The shopper who bought nothing, and how a survey reaches them
Look down the middle column. A receipt code, a link in an order email and a loyalty list each need a purchase before they can reach anybody at all.
| Where you put the link | Who it reaches | Who it cannot reach | What the answers are good for |
|---|---|---|---|
| A code on the receipt, or a link in the order email | Whoever paid, once they have paid. | Everybody who walked out without buying, and anybody who paid cash and left the receipt. | The checkout and whatever happened after it. |
| A tablet or a code at the register | People still at the counter, while the visit is in front of them. | Anybody who never joined a queue, which on a browsing visit is the whole point. | The till, at the busiest minute of the visit. |
| A code on the door, a shelf card, a fitting-room mirror or a bag insert | Anyone who came in, whether or not they bought anything. | People who looked at the window and kept walking. | The whole visit, with buyers and non-buyers in one set you can split. |
| An email to your list | Customers who already gave you an address: an account, a loyalty card, a previous order. | Every walk-in, who by definition gave you nothing to write to. | Buyers you already have an address for, and nothing at all about passing trade. |
Only the third row can reach a person who came in and left with nothing, and that person is often the one with the most useful thing to say: they wanted something enough to walk in, and something in the store sent them out again. Question 9 is written for them, and the link has to be somewhere they will pass on the way out.
The published methods all sit at the register. A guide to running these in a store describes tablets and kiosks placed at the cash register, and codes printed on physical media, as the usual approaches.[2] A kiosk by the register is answered by somebody who has already decided to pay. A code on the door, a shelf card, the mirror in a fitting room or a slip in the bag costs the same to print and does not have that condition attached.
The store is still where most of the money is spent. The Census Bureau measured e-commerce at 17.1 per cent of total US retail sales in the second quarter of 2026.[1] That is a measurement of sales rather than of visits, and it is revised every quarter, so take it for the one thing it says and no more.
What a printed code cannot do. It resolves to an ordinary survey link, which is why it needs nothing special to work. It also cannot tell which of your stores the person was standing in, and it cannot stop one person answering twice. The first of those is what question 1 exists for. The second is why a single strange answer is worth reading rather than acting on.
A paid observer is a different survey. If what you want is a trained shopper working through a script and scoring the store against a checklist, that is the mystery shopping survey template: the person answering it came in to do a job rather than to buy something, and their visit is the one visit in the building that was never going to end in a purchase.
Reading twelve answers from one store visit
The order you read them in is what turns twelve ratings into a decision about one store.
Split on question 2 before you read anything else. A buyer and a non-buyer answered the same twelve questions about two different visits. For somebody who bought, a low answer on question 3 is a detail on the way to a sale. For somebody who left with nothing, question 3 is the entire visit, and question 9 is the sentence the survey was printed for.
Then split on question 1. Reading every location together produces one number that describes no store you own. Zoho's own guidance to a multi-site retailer is to filter responses by store, region or time period rather than read the aggregate.[3] It sells that as a feature of a reporting product; here it is a question you ask, which is why it comes first and why it cannot be added afterwards.
Read the close against the block above it. Question 12 asks whether the visit went the way they wanted. A high answer sitting over a weak middle is not a contradiction to explain away: it says somebody forgave you. Note whose store it was, and go back to whichever item was weak.
One visit is not a journey. If the subject is how somebody found you, decided, bought and came back, the customer journey survey questions cover six stages in one send and name the weakest one.
Nor is it an order. When the thing being rated starts at a checkout and ends when a box arrives, the post-purchase survey template splits that into separate sends around one order.
A guest is not a walk-in. Somebody booked in, hosted and known by name before they arrive is a different respondent with a different survey, and the guest experience survey is built for them.
And a supermarket is its own thing. When the store is a supermarket and the visit is the weekly shop, the grocery store survey questions covers the departments and the substitutions this one leaves out.
If you would rather not use ratings at all, and want open answers in one pile with themes pulled out of them, the customer feedback questionnaire is built that way from the start.
Retail customer survey FAQ
What is a retail customer experience survey?
A short questionnaire about one visit to one store, answered by the person who made the visit. This one has twelve questions: which store, whether they bought anything, six ratings covering the errand, the staff and the state of the place, two open boxes and two ratings to finish. It takes about two minutes on a phone.
How do you survey somebody who left the store without buying?
Put the link somewhere that is not the till. A code on the door, a shelf card or the mirror in a fitting room can be scanned by anyone who came in, whether or not they reached a register. A receipt code and an order email cannot, because both need a purchase before they exist. Question 9 is written for that person.
How many questions should a retail survey have?
Twelve here, answered in about two minutes. That is a design choice rather than a finding, and there is no published number to appeal to, so treat any recommended count with suspicion, this one included. The constraint that actually binds is where it gets answered: on a phone, in a car park or on a bus, minutes after walking out.
How do you compare results across store locations?
Ask which store first, as question 1, and read each store on its own rather than pooled. Rename the four options to your own stores before you send it. There is no way to add that question back once the answers are in, which is the reason it goes first rather than last.
What can you use instead of a paper feedback form in a store?
A printed code that opens the survey on the shopper's own phone. It costs nothing to reprint when a question changes, it does not need anybody to carry a card across the store to a box by the door, and the answers arrive typed rather than handwritten. Keep one paper form at the counter for anyone who would rather write.
Methods and sources
What this template is based on
The question order follows the shape of a visit: which store, whether it ended in a purchase, the errand, the people, the place, the till, then their own words. The channel table lists what each way of asking can physically reach, which is a property of the channel and not a finding. The share of US retail sales that is e-commerce is the Census Bureau's own quarterly measurement[1], and the description of how in-store surveys are usually run is taken from a published vendor guide[2].
Question licensing
Nothing licensed is reproduced. No proprietary questionnaire sits behind a retail store survey, so every stem here can be copied, reworded and sent with no attribution to anybody, including us.
Limits and disclosure
The Census figure measures sales and not visits, covers the United States, and is revised every quarter, so it is quoted with its quarter and its release date attached.[1] The two vendor pages are cited for their description of method only; the statistics printed alongside them there are second-hand and none of them is used here.[2][3] The widely quoted figure for how many shoppers leave a store without buying traces back to footfall vendors with no published method, so this page makes the argument without a number and you should take yours off your own door counter. No target score is published here either, because a retail score is built on your stores, your questions and where you put the link.
SuperSurvey makes the editor this template opens in. Everything on this page works with or without it.
References
- US Census Bureau. Quarterly Retail E-Commerce Sales, 2nd Quarter 2026. Released 18 August 2026: "E-commerce sales in the second quarter of 2026 accounted for 17.1 percent of total sales." census.gov
- QuestionPro. In-store surveys: types, benefits and guide with example. Cited for its description of method only. questionpro.com
- Zoho Survey. Retail customer experience surveys. Cited for its instruction to filter responses by store, region or time period rather than read an aggregate. zoho.com
Put the code by the door, not by the till
Twelve questions, about two minutes, and a second question that separates the people who bought from the people who did not. Open it in the editor, rename the store options and print one link.
Use this template