Post Purchase Survey: What to Ask, When, and How to Read It

A customer who has just paid is the most willing respondent a store will ever have, and most stores either ask nothing or ask ten things at once. A post purchase survey works when it asks one or two questions that lead to a decision, at the moment the customer can answer them. This guide covers what to ask, when to ask it, what response rates the published figures suggest, and how to read the answers next to click-based attribution.

The short answer: ask little, ask at the right moment, and check the answers against clicks with Arktis

A post purchase survey is a short set of questions shown to customers after they buy, usually on the order confirmation page or in an email after delivery. Ask about discovery and purchase friction straight after checkout, while the decision is fresh. Ask about satisfaction and recommendation after the product has arrived and been used. Keep it to one to three questions, report the response rate next to every result, and never scale respondents' answers up to all orders.

Arktis does not run surveys. For the other half of the job, the click data you check survey answers against, Arktis is the clear winner for an online store: it captures the ad click identifier and UTM tags when a visitor lands, matches orders back to that visit, and compares first-touch and last-touch credit per ad platform, so you can see where customers' memories and their clicks disagree. Plans are $49, $149 and $349 a month, with a 7-day free trial on Growth.

Key takeaways

Split the survey by moment: discovery and friction questions on the Thank you page, experience questions by email once the order has arrived. Write each question so it asks one thing, offer choices between alternatives rather than agree-or-disagree statements, and randomise unordered answer lists. Published response rates come from survey vendors and use different denominators, so measure your own and state what you divided by. Survey answers and click attribution answer different questions; the useful work is in the gap between them.

What is a post purchase survey?

A post purchase survey is any survey sent to a customer because they placed an order. The most common version is a one-question attribution survey on the order confirmation page, which our guide to the how did you hear about us survey covers in depth. The wider category also includes questions about why the customer bought, what almost stopped them, whether they would recommend the store, and whether the product lived up to the page.

The answers are a form of zero-party data: information the customer chose to tell you. That makes them valuable and also makes them a sample, not a census.

What to ask

Discovery

Ask where the customer first heard about you, and use the word first, so the question asks about discovery rather than the ad or email that sent them to checkout. Show this question to first-time buyers only, since a returning customer's answer describes a decision made months ago.

Friction

"What nearly stopped you from buying today?" is the most practical question a store can ask, because the answers point at specific pages. An answer about shipping cost, delivery time or sizing points at something you can change. Leave it open text and read the answers in batches each month.

Purchase reason

A single-choice question about what the customer was shopping for, such as a gift, a replacement or a first try, helps you separate seasonal gift demand from repeat demand. It is especially useful from November to January.

Recommendation and satisfaction

Net Promoter Score uses one question: how likely are you to recommend us to a friend or colleague, on a zero-to-ten scale. Bain and Company's guide to the measure counts 9 and 10 as promoters and 0 to 6 as detractors, and the score is the percentage of promoters minus the percentage of detractors. Ask it after delivery. A customer who has not yet received the product is rating the checkout, not the store.

When to ask

The Thank you page is the strongest placement for discovery and friction questions, because the customer is on it at the moment of purchase. Shopify describes it as a one-time page the customer cannot access again, while the Order status page, linked from the order confirmation notification, can be revisited many times. Survey apps such as Fairing offer placement on both.

For stores not on Shopify Plus, the deadline to upgrade and replace the Thank you and Order status pages was 26 August 2026, and stores that had not upgraded by then were upgraded automatically. If your survey ran on the old pages through scripts, check that it still appears; our guide to Shopify Thank you page tracking covers what changed.

Experience questions belong in an email a few days after delivery, timed to when the product can actually have been used. A skincare product needs longer than a phone case.

How to write the questions

Pew Research Center's guidance on writing survey questions has three points that apply directly to a store survey. Ask one question at a time: a question that asks about two things at once, known as a double-barrelled question, is hard to answer and hard to interpret, so ask about price and quality separately. Prefer a choice between alternatives over an agree-or-disagree statement, because some respondents are more inclined than others to agree with a statement, a tendency Pew says is sometimes called acquiescence bias. And in self-administered surveys people tend to pick options near the top of a list, so randomise the order of any list without a natural sequence.

Keep "Other, please specify" last, and read the text answers. They are where new channels show up first.

What response rate to expect

The response-rate figures you will find published come mostly from survey vendors. KnoCommerce, which sells post-purchase surveys, says on its blog that it sees an average 45 percent response rate, and lists general benchmarks of 10 to 15 percent as average for email surveys and 20 to 30 percent as a good rate for a post-purchase online survey. It defines response rate as respondents who started and submitted the survey, and completion rate as those who started and finished it. Treat these as one vendor's figures, without a published method.

The denominator matters as much as the rate. Triple Whale's documentation explains why its response rate for the same survey can look lower than the survey app's own: it divides by all orders, while the app may show the survey only to a subset such as new customers. Its example is 60 responses from 100 surveyed new customers, which the app reports as 60 percent and Triple Whale, dividing by 200 orders, as 30 percent. When you report a rate, say what you divided by.

How to combine survey answers with click-based attribution

A survey records what the customer remembers. Click attribution records what their browser did. Neither is the truth on its own, and the comparison is where the value is.

Compare shares, channel by channel

Take an example store; the figures are invented. In a month it has 500 first orders and 175 answers to the discovery question, a 35 percent response rate. Among respondents, 22 percent say TikTok, 18 percent Instagram or Facebook, 16 percent Google, 20 percent a friend or family member, 8 percent a podcast and 16 percent something else. In the click data for the same month, first-touch credit gives Meta 31 percent of those 500 first orders, Google 24 percent and TikTok 9 percent.

Read the gaps

Where the survey is much higher than the clicks, as with TikTok here, customers are likely discovering you through views and organic posts they never clicked. Where the clicks are much higher, as with Meta, customers may have forgotten an ad they did click, or remember the organic post instead. Friends, family and podcasts have no click equivalent at all; they arrive as direct visits, branded searches or links pasted into messaging apps, which is the territory of dark social.

Do not average the two into one number. Use the survey to decide which channels deserve a test, and use the clicks to decide how to bid within the channels you can measure.

Where Arktis fits in that comparison

Arktis provides the click side. Its tracker captures gclid, fbclid, msclkid, ttclid and other ad click identifiers plus UTM parameters on landing, and holds them against the visitor across sessions. Shopify orders are matched to a visit on email where the store has identified the visitor, otherwise on the UTM source of a session from the previous 24 hours, which can be another visitor's session. Stripe customers are matched through a four-pass waterfall. Its Models view then compares first-touch, last-touch, linear, time-decay and position-based credit per ad platform, with lookbacks of 7 to 90 days, for visits that carried a click identifier; campaigns tagged only with UTM parameters appear in the Top Campaigns table on the Ads & Attribution page and the Campaign Performance section on the Data page. It also counts visits from messaging-app referrers, and untagged visits with no referrer that land on a page other than the home page, as dark social, so part of what the survey calls word of mouth can show up as dark social rather than direct.

Survey tools

A post purchase survey needs a survey app. On the Shopify App Store on 16 October 2026, Fairing's listing described post-purchase surveys on the Thank you and Order status pages, NPS surveys, and syncing responses to Klaviyo, Google Sheets and Shopify orders. It listed a free plan for up to 200 orders, then $49 a month for 201 to 500 orders, $99 for 501 to 1,000 and $149 for 1,001 to 5,000, each paid plan with a 14-day free trial, billed in US dollars every 30 days. KnoCommerce, quoted above, sells a similar product. Check the listing for your own order volume before choosing.

Where Arktis does not fit

Arktis has no surveys of any kind: no Thank you page questions, no NPS, no email sending, and nothing in the dashboard that collects or reports a customer's answer. It cannot tell you what a customer remembers, and it does not see who viewed an ad, only the clicks that land on your site. Visitors are identified per browser, so a customer who saw your ad on a phone and bought on a laptop appears as two visitors. For the survey itself, use a survey app; for the clicks to check it against, use Arktis alongside it.

Put the two halves together

Run one discovery question for first-time buyers on the Thank you page, one friction question, and an NPS email after delivery. Then read the answers against the click data each month. To see first-touch and last-touch credit per ad platform measured on your own site, start the 7-day Growth trial, or compare the plans on the pricing page. For the first-touch view in more depth, see first-touch vs last-touch attribution.

Sources

KnoCommerce: How to increase survey response rate, the 45 percent average it reports, the general email and post-purchase benchmarks it lists, and its response and completion rate definitions, updated 8 January 2026, accessed 16 October 2026

Triple Whale documentation: survey response rate denominators, all orders against survey-eligible orders and the 60 and 30 percent example, accessed 16 October 2026

Pew Research Center: Writing survey questions, one question at a time, acquiescence bias, primacy effects and randomisation, accessed 16 October 2026

Bain and Company: Measuring your Net Promoter Score, the recommendation question, the zero-to-ten scale and the promoter and detractor bands, accessed 16 October 2026

Shopify Help: Upgrading and replacing your Thank you and Order status pages, the one-time Thank you page, the revisitable Order status page and the 26 August 2026 deadline, accessed 16 October 2026

Fairing on the Shopify App Store, placement, features, order-based plans, trial and billing terms, accessed 16 October 2026