By 
A post purchase survey is the highest-return question in ecommerce, and most stores waste it. The moment after checkout is the one time a customer is paying full attention, has just trusted you with their money, and is happy to tell you why. Most stores fill that moment with five questions they will never act on, or with the wrong question entirely.
This is a guide to getting it right. What a post-purchase survey is for, the exact questions worth asking, a template you can copy this afternoon, when and where to run it, and the one thing it will never be able to tell you no matter how well you write it.
First, name the job
Before you write a single question, decide what the survey is for. Shopify stores run post-purchase surveys for two genuinely different jobs, and the question stack that wins at one is close to useless at the other.
The first job is attribution: where did this customer actually come from? This matters more than it used to. Since Apple’s App Tracking Transparency arrived with iOS 14.5 in April 2021, a large share of iPhone users opt out of the tracking that ad pixels rely on, so your platform analytics increasingly cannot tell you where someone first heard of you. Asking the buyer directly, on the thank-you page, is now one of the few honest reads on channel performance you can get.
The second job is satisfaction: how did the buying feel, and would they do it again? This is your NPS, your CSAT and your product feedback, asked while the experience is fresh. It is a voice-of-customer exercise run at the best possible moment.
These jobs pull in opposite directions on timing and wording, which is why trying to do both in one survey usually does neither well. Attribution wants the answer the instant the order confirms. Satisfaction, at least the part about the product and the delivery, wants a few days for the parcel to actually arrive. Name the job first. The rest of this guide is organised around that split.
The best post-purchase survey questions
Good questions share one trait: the answer changes something you do. If you already know the answer from your analytics, or you would not act on it either way, cut the question. Here are the ones that earn their place, by job.
Attribution questions
“How did you hear about us?” The single highest-value question most stores can ask. Keep it open text, or offer a short list of channels with an “other” box. If you use a list, rotate the order so the top option does not soak up lazy clicks.
“Was there a particular reason you bought today?” This is the tipping-point question. Attribution tells you the channel; this tells you the trigger, the review, the discount, the out-of-stock alert, the friend’s recommendation. It is the difference between knowing someone saw your TikTok and knowing what in the TikTok made them buy.
“Is this your first order with us?” Optional, and only if your order data does not already answer it cleanly. Useful for splitting new-customer attribution from repeat behaviour.
Satisfaction questions
The NPS question: “How likely are you to recommend us to a friend or colleague?” on a 0 to 10 scale. This is the metric Fred Reichheld introduced in his 2003 Harvard Business Review article, The One Number You Need to Grow, and it has stuck around for a reason: the willingness to put your own reputation on the line by recommending a brand tracks loyalty better than a bland satisfaction rating. Scores of 9 and 10 are promoters, 7 and 8 passives, 0 to 6 detractors, and your Net Promoter Score is the percentage of promoters minus the percentage of detractors.
“What is the main reason for your score?” open text. The score is a thermometer. This is the diagnosis. Never ask the NPS number without this follow-up, because the number on its own tells you the temperature and nothing about the illness.
“How did the product compare to what you expected?” Optional, and best asked a few days post-delivery. It surfaces the gap between your marketing promise and the real thing, which is where returns and one-star reviews come from.
Questions to cut
Some questions show up on every listicle and earn their keep on almost none.
Anything your analytics already answers. You know the order value, the products bought, the device and the location. Do not spend a precious question slot re-collecting data you already hold.
Double-barrelled questions. “How satisfied were you with the price and delivery?” cannot be answered cleanly, because the customer may have loved one and hated the other. Ask one thing at a time.
Leading questions. “How much did you love our new packaging?” tells you nothing except that you like your own packaging. Ask neutral questions or you will collect flattery.
Long demographic grids. Age, income, household size, all the market-research furniture. Every extra question costs you responses, and on a post-purchase survey you are spending goodwill you will want later.
A post-purchase survey template you can copy
Here are two ready-to-run stacks. Pick the one that matches your job, keep it to the questions shown, and resist the urge to add a fourth.
Attribution stack, on the thank-you or order-status page:
- How did you hear about us? (open text, or a rotated channel list with an “other” box)
- Was there a particular reason you bought today? (open text)
- (Optional) Is this your first order with us? (yes / no)
Satisfaction stack, in an email two to three days after delivery:
- How likely are you to recommend us to a friend or colleague? (0 to 10)
- What is the main reason for your score? (open text)
- (Optional) How did the product compare to what you expected? (better / about the same / worse, with an optional comment)
That is the whole thing. Two required questions and one optional per survey. If you run both stacks, you have covered the two jobs a post-purchase survey does well, with a total of four required questions across two touchpoints, and you have not tested your customers’ patience once.
When to send it, and where
Timing is not a detail. It is most of the result.
Ask attribution questions immediately, on the thank-you page or the order-status page, the instant the order confirms. The decision is fresh, the customer is still in the flow, and the friction of answering is close to zero because they are already looking at the screen. Wait a week and they genuinely will not remember whether it was the email or the Instagram ad that tipped them over.
Ask product and delivery questions a few days after the parcel arrives, by email or on the order-status page. There is no point asking how the product compared to expectations before the box has landed. For these, a short email survey timed to roughly a day after delivery does the job. The order-status page is quietly the best home for the delivery question, because customers reopen it again and again to track the parcel, so a question sitting there gets seen at exactly the moment the box turns up, without you sending a thing.
The general rule: ask about the checkout in the checkout, and ask about the product once the product exists in the customer’s hands. Match the moment to the question and your answers get sharper and your response rate climbs.
How long should it be?
Short. Shorter than you think, and then one question shorter than that.
The data here is unambiguous. Survicate’s 2025 benchmark, drawn from 4,332 surveys across 460 companies, found that microsurveys of two to three questions had the highest median response rate at about 16 per cent, comfortably beating single-question surveys at around 10 per cent and surveys of seven or more questions, which dropped to under 7 per cent. Every question you add past the third is costing you responses, and a survey nobody finishes is worse than no survey, because it flatters you with a biased sample of the few who had time.
The same dataset puts the overall median response rate for on-site and in-app surveys at just under 10 per cent, with a median completion time of 42 seconds. That is your budget. Roughly a minute of a customer’s attention and, on a good day, one in six of them saying yes. Spend it on the two or three questions that will change a decision, and nothing else.
What a good response rate looks like
Set your expectations with real numbers, not a vanity figure.
A well-placed, short post-purchase survey on the thank-you page, where attention is highest, will typically sit at the upper end of the Survicate benchmark range: think mid-teens as a percentage, with the strongest performers pushing past 20 per cent. Retail and wholesale as a category ran a median of about 14 per cent in that dataset, ahead of the 10 per cent all-industry median, which fits with consumers being more willing to answer than business buyers. An emailed survey sent days later will land lower, usually somewhere in the low double digits, because you have lost the peak-attention moment.
Two honest caveats. First, response rate is not representativeness. The people who answer skew towards the delighted and the furious, so treat the quiet middle as under-represented in your results. Second, chase your own trend rather than a magic number. A survey that climbs from 8 to 14 per cent because you shortened it and moved it to the thank-you page is worth far more than agonising over whether 14 is a “good” figure in the abstract.
And think twice before buying that response rate with an incentive. A discount or loyalty points will lift the number, but never dangle one in front of an attribution survey: you will pull in people answering for the reward rather than the truth, and the “why did you buy” answer is the first thing to rot. A light incentive is more defensible on a satisfaction email, as long as you remember it nudges NPS up and read the trend accordingly. The honest response rate you get for free is worth more than the flattering one you paid for.
What to actually do with the answers
Collecting the data is the easy part. Most stores stop there, and the survey quietly becomes a dashboard nobody opens. Here is how to make the answers pay.
Reconcile attribution, do not obey it. Self-reported attribution is directional, not precise. People name the last thing they remember, which is rarely the first touch that actually started the journey. So do not reallocate your budget on the raw survey alone. Put the survey’s channel breakdown next to your platform analytics and look for agreement and disagreement. Where both say a channel is pulling its weight, spend with confidence. Where they diverge sharply, that is a flag to investigate, not a number to act on blind. The survey is one witness, not the verdict.
Read the NPS verbatims, not just the score. The number is a trend line to watch quarter on quarter. The open-text reasons are where the actual work lives. Tag them into themes, delivery, sizing, product quality, support, and you have a ranked list of what to fix, sourced straight from the people who just paid you.
Close the loop on detractors. A customer who scores you a 3 and tells you why has handed you a chance to fix it before they write the review or request the refund. Route low scores to a human. This is the highest-return follow-up in the whole exercise, and almost nobody does it.
Feed the tipping-point answers to your marketing. When customers keep naming the same review, the same creator, the same objection you finally overcame, that is your next ad, your next landing-page headline, your next FAQ. The “why did you buy” question is a free stream of copy tested by people who actually converted.
The one thing a post-purchase survey can’t tell you
Every question above shares one hard limit, and it is worth being clear-eyed about it. A post-purchase survey can only ask people who already bought. By definition, it is a read on the past.
That is exactly what you want when the question is where a customer came from or how the checkout felt. It is no help at all when the question is whether something will sell in the first place. And most of your demand never reaches the survey. Baymard Institute’s aggregate of dozens of studies puts the average documented cart-abandonment rate at around 70 per cent, and Littledata’s Shopify benchmark puts the average store’s conversion rate at roughly 1.4 per cent. So a post-purchase survey is, in effect, interviewing the 1 or 2 per cent who said yes and never hearing from the 98 who did not. It cannot tell you why the others left, and it certainly cannot tell you whether a product, pack, price or claim you have not launched yet is worth putting into the world, because there is nobody to survey. The thing does not exist.
That gap is a different job, and it belongs before you spend, not after. It means testing your audience assumptions on the idea itself, ahead of committing stock or ad budget. This is the one place our own tool fits, so I will name it plainly. With TestFeed you can put a product, a pack, an in-context price, an ad or a claim in front of simulated shoppers and get back a purchase-intent read, their reasons in their own words, and a clear next move, in days rather than weeks. It is a pre-spend, directional signal, not a sales forecast and not a substitute for asking real buyers once you have them, and it does not judge taste, texture or smell. Use it to decide which idea is worth launching, then point a post-purchase survey at the customers it brings you. One looks forward before you spend. The other looks back once you have.
Which app should you run it on?
You do not need a dedicated tool to start. Shopify’s own thank-you and order-status pages can carry a simple question, and most email platforms will send an NPS follow-up. When you outgrow that, the specialist apps split roughly by job. Fairing and KnoCommerce lead on attribution and join answers to order and lifetime-value data. Zigpoll is the flexible all-rounder across post-purchase and on-site. Grapevine is the flat-rate value pick for high-volume stores. Prices move often, so confirm on each app’s Shopify App Store listing before you commit, and pick by the job you named at the top of this guide rather than the star rating.
Frequently asked questions
What is a post-purchase survey?
A short survey shown to a customer straight after they buy, usually on the thank-you or order-status page or in an email a day or two later. It exists to answer one of two questions well: where the customer came from (attribution) or how the purchase felt (satisfaction). Because it catches people at peak attention, it gets far higher response rates than a survey emailed cold days later.
What are the best post-purchase survey questions?
The ones tied to a decision you will actually make. For attribution, ‘How did you hear about us?’ plus ‘Was there a particular reason you bought today?’. For satisfaction, an NPS or CSAT score followed by an open ‘What is the main reason for your score?’. Skip anything your analytics already tells you, and skip double-barrelled or leading questions.
When is the best time to send a post-purchase survey?
It depends on the job. Ask attribution questions immediately, on the thank-you or order-status page, while the decision is fresh and the response rate is highest. Ask delivery and product-satisfaction questions a few days after the parcel arrives, so the customer has actually used the thing you are asking about.
How many questions should a post-purchase survey have?
Two or three. Survicate’s 2025 benchmark of 4,332 surveys found two-to-three question surveys get the highest median response rate, about 16 per cent, while surveys of seven or more questions drop to under 7 per cent. Every extra question costs you responses, so make each one earn its place.
Can a post-purchase survey tell me if a new product will sell?
No. A survey can only ask people who already bought or visited, so it describes the past. To gauge demand for a product, pack, price or claim you have not launched, you need pre-launch testing, which puts the concept in front of a modelled audience before you commit stock or ad spend. That is a different job from any post-purchase survey.
The survey to run this afternoon
Put one attribution question on your thank-you page: how did you hear about us, with a “why did you buy today” follow-up. Send one NPS question by email two days after delivery, with an open reason box. Keep each to two or three questions. Reconcile the attribution answers against your analytics before you move a penny of budget, read the NPS reasons every week, and route every low score to a real person. Do that and you will learn more from the customers you already have than most stores learn from a full research agency, at the cost of about a minute of each buyer’s time.
