Customer Experience Survey Questions and Template
One survey, sent once, to people who have recently been all the way through: sixteen questions across finding you, buying, getting going, living with it, getting help and staying. It does not score the relationship. It names the stage that has earned a permanent survey. Answer it below, then copy the questions or open it in the editor.
- 16questions
- About 4 minto complete
- One sendnot a repeating wave
- A stagenot a score
Which stage to measure next, and the survey that belongs there
Every published list of customer experience survey questions hands over six stages and leaves the reader to guess which one to instrument. This table goes the other way. Start at the stage you suspect, or take the top row.
| The stage you suspect | What that looks like in the answers | The survey that belongs there |
|---|---|---|
| No stage named yet, which is the normal starting point | Nothing stands out, or nothing has been measured at any stage. | The 16-question audit on this page, sent once |
| Finding you, and deciding you were worth trying | The first two questions sit below the rest of the form. | Add a brand awareness survey to the market |
| The buying moment itself | Questions 3 and 4 drag while everything after them holds up. | Put a scored customer satisfaction survey on it |
| Living with it, month after month | The early stages read well and the middle of the survey sags. | Run the quarterly voice of the customer survey |
| The time something went wrong | Few people answer questions 9 and 10, and those few mark them lowest. | Fire a customer service survey for one request |
| Coming back, or naming you to somebody else | The last block falls away from the rest, whatever the earlier stages say. | Stand up the NPS survey and its calculator |
| Every stage, but in sentences rather than scores | The written boxes fill up and the ratings barely move. | Open an open-text customer feedback survey |
The table reads both ways. Going in, the first column is a shortlist of what you already suspect. Coming back out, the middle column is a lookup: find the pattern your results made, and that row names the survey the stage keeps.
The published advice is to survey each stage separately, three to five questions at a time.[2][3] Good advice, one step further along than most readers are, because it assumes the stage has been chosen. Rawson, Duncan and Jones argued in Harvard Business Review that a narrow focus on maximising satisfaction at individual touchpoints can create a distorted picture, suggesting that customers are happier with a company than they actually are.[1] This is the cheap way to test it.
The 16 customer experience survey questions
Each question, the scale it uses, and what the answer says about that stage. Copy one stage, copy the lot with the store swap-in, or open the survey and skip the typing.
Finding you
Two statements about the stretch before any money changed hands.
- When I was looking, it was clear what you actually offered.Whether the offer explained itself. A low answer from somebody who bought anyway means a person had to do the explaining.
- By the time I decided to go ahead, I knew what I would be getting.The picture the customer carried into the purchase. Disagreement means the description and the thing came from different understandings.
Buying
Two statements about the transaction, kept apart from anything that happened after it.
- Buying from you was straightforward once I had decided.Friction in the checkout, the form or the contract. This item is the process; the next is the price.
- The price I paid was the price I had been shown.Charges, add-ons and small print, which turn up again in the written box in their own words.
Getting going
Two statements about the first days, which decide whether the rest gets used at all.
- Getting started took about as long as I thought it would.Setup measured against expectation, not against a clock. A slow start nobody was warned about scores worst of all.
- I could work out how to get going without asking anyone.Self-service. Disagreement here, with agreement in the help block, is a support team paying for a product problem.
Living with it
Two statements about the long middle, where a customer spends most of their time and gets least of your attention.
- Since then, it has done the job I got it for.Basic fit, months in. Disagreement from somebody still paying you means they have quietly worked around you.
- It has become easier to live with, not harder.Direction of travel: a product being looked after against one that shipped well and then stopped.
If something went wrong
Two optional statements, because they apply only to the people something went wrong for.
- When something went wrong, I could reach a person who could help.Reachability: getting to a person at all, not how that conversation went.
- The problem I had was put right without me having to chase it.Chasing. A low answer means the fix arrived and the customer managed it.
Staying, or not
Two statements about whether the relationship survives, neither of them the recommend question.
- Nothing has happened since that made me look at what else is out there.Whether anything pushed them to look elsewhere. Disagreement is a customer halfway out and still paying.
- If I were making this decision again today, I would end up in the same place.The whole journey judged in the one decision they have already made. The closest this gets to a verdict.
Name the stage, and the words for it
The nomination item and the two open boxes. Question 13 is what makes this an audit, not a list.
- Which part of this was the worst one for you?The direct ask. A stage the ratings rank third and the customers name first is the one to fix.
- What is the one thing you would change about how this went?Where a problem sitting underneath every stage, like price, shows itself.
- What went better than you thought it would?A stage that works is worth protecting once fixing starts, and nothing else here finds one.
- Taking all of it together, this went the way I hoped it would.A high close over a weak block is a customer who forgave you, which is a different problem.
Swap-in: one store visit
For a journey that happened in a shop: replace questions 1 to 6 with these four. When one visit is the whole subject, a grocery store feedback survey fits better.
- I could find what I came in for.Findability, which in a shop stands in for the whole finding-you stage.
- Somebody was around when I needed one.Staffing as the customer met it, not as the rota describes it.
- Paying took about as long as I thought it would.The queue, judged against expectation rather than in minutes.
- The place looked the way I expect it to look.Standards, first to slip when a store is short-staffed.
Three questions are missing on purpose. No 0 to 10 recommend item, no rating of how satisfied anyone was with a single experience, no question asking how hard anything was. Each is a whole survey in the table above, and each stops working buried at question nine of sixteen. Nobody is asked to count anything either: a person answering about a finished journey builds the answer from partial memory and inference.[4]
Why sixteen questions and not fifty
Two questions per stage, six stages, nothing spare. The first item in each pair asks whether the stage worked, the second whether it worked the way the customer had been led to expect. Enough to rank six stages against each other, not enough to describe any one, which is the right trade for a survey whose output is a name.
One scale, so the blocks can be compared. Rate one stage out of ten, another out of five and a third with a smiley face, and you get six numbers that cannot be set side by side. Question 13 then asks the customer to do the ranking too, and when the two disagree, instrument the stage the customers named.
The published lists run past fifty for a reason that is not yours: a page can afford fifty questions because nobody sends a page. All sixteen here are already built into the survey maker online, so change what does not fit, drop a stage you do not have, and send it.
Sending the journey audit once
Five decisions that belong to a survey with an end. What every send has in common is in the survey research guide.
Pick people who reached the end, recently
Everyone on the list has to have gone all the way through, inside the last month or two. Somebody halfway along cannot answer about stages they never reached, and their blanks look like a stage that failed.
Name the journey in the first line of the invitation
Write the actual thing: the order placed in March, the plan signed up for, the job done at their address. An invitation about "your experience with us" gets answered about your company.
Ask about one journey, never about a habit
No question here asks how many times anything happened, and none should be added. Somebody reporting on a finished stretch of their life rebuilds it rather than recalls it, so a count is invented the moment you ask.[4]
Leave the help pair skippable
Questions 9 and 10 apply only to the people something went wrong for. Marking them optional keeps everyone else out, and the share who answer anyway reads how often the journey breaks.
Close it, choose one stage, and stop
Take the single worst stage, put the permanent survey named in the table on that one, and leave the other five alone until it has told you something.
Turning the answers into one stage
Read it in three passes. The nomination counts from question 13 first, because that is the customer choosing. Then the six blocks, averaged inside each block and never across them. Then the written boxes. A weak block under a high close is a stage people forgave you for; under a low close it is one they still hold against you.
The honest limit, which this survey cannot design its way out of. A stage somebody passed months ago is not so much recalled as rebuilt, out of partial memory and inference, and what gets built leans towards how the person feels about you on the day they answer.[4] That literature is health research rather than a study of customers, so treat it as a warning and not a measurement. It is why this survey points at a stage instead of scoring one: a ranking survives that error far better than a number does.
What it will not give you. Not a percentage, not a trend, not the size of the shortfall at the stage it names. If what you want is the SERVQUAL gap between expectation and perception across five named service dimensions, that is a different instrument, and it belongs on one stage.
Customer experience survey FAQ
What is a customer experience survey?
A survey about the run of stages one customer went through with you, rather than one moment inside it. Send it once, to people who have recently been all the way through, and read it for the stage that came out worst.
How many customer experience survey questions should you ask?
Sixteen here, and three to five in whatever you leave running at the stage this one names.[2]
Who should get it, and when?
People who have been all the way through recently: found you, bought, used it, and either stayed or did not. Not prospects, not somebody three days in, and not the whole list. The nearer the send sits to the end of the journey, the less of it is rebuilt from memory.[4]
Can these customer journey survey questions be split into shorter surveys?
Yes. Each pair is the seed of a permanent survey at that stage, and the table above names the template that covers it. Starting all six at once gives you six things to maintain and no way to rank them.
What if the answers do not point at one stage?
Then read the worst-part question and the written boxes first. Flat blocks under a low close say the trouble is not at a stage at all: price, or the thing itself, sits underneath all six.
The stage axis on this page is deliberately generic. When the whole journey happens inside one shop on one afternoon, the retail survey is scoped to that single visit instead, including a visit that ended without a purchase.
Methods and sources
What this template is based on
The stage structure and the length rule quoted for a stage survey follow the published convention of surveying each journey stage separately[2][3], the case for measuring the run of stages rather than the moments inside it is attributed to its authors[1], and the consequences of asking about a finished journey follow the literature on retrospective self-report[4].
Question licensing
Every question here is ours, and free to copy, edit and send anywhere. Net Promoter, NPS and Net Promoter Score are marks of Bain and Company, NICE Systems and Fred Reichheld; the recommend question is named in the routing table and never reproduced.[5]
Limits and disclosure
No benchmark figure is offered. The recall caution comes from health research and has not been measured on customers[4], and the case for the journey over the touchpoint is a published position rather than a finding this page can verify[1]. The survey ranks six stages; it does not size the gap at any of them.
SuperSurvey makes the editor this template opens in. The sixteen questions are free to use with no account.
References
- Rawson, A., Duncan, E. and Jones, C. The Truth About Customer Experience. Harvard Business Review, September 2013. hbr.org
- Formbricks. Customer Experience Survey Questions by Journey Stage. Formbricks blog. formbricks.com
- Attest. Customer survey questions for every step of the customer journey. Attest blog. askattest.com
- Schwarz, N. Retrospective and Concurrent Self-Reports: The Rationale for Real-Time Data Capture. In A. Stone, S. Shiffman, A. Atienza and L. Nebeling (eds.), The Science of Real-Time Data Capture, pp. 11 to 26. Oxford University Press, 2007. dornsife.usc.edu
- Bain and Company. Measuring Your Net Promoter Score. Net Promoter System. netpromotersystem.com
Find the stage before you buy a survey for it
Sixteen questions, about four minutes, sent once to people who have been all the way through. Open it in the editor, change any wording, and pick your list.
Use this template