How did you hear about us? Post-purchase surveys as your cheapest attribution layer

One question on your thank-you page tells you more about your channels and creatives than most dashboards. Here is how to use post-purchase surveys as a triangulation layer.

A post-purchase survey is one simple question on your thank-you page: how did you hear about us? It is the cheapest way to check what your attribution is telling you. Meta claims purchases, Google claims purchases, and your customer is the only one who knows what actually happened. For the price of an app or a simple form, you get a data layer that sharpens your channel decisions and your creative decisions more than most dashboards ever will.

What exactly is a post-purchase survey?

Right after checkout, on the thank-you page, you ask the buyer one or two short questions. Not a long questionnaire, not an email three days later, but a question at the exact moment the purchase is fresh and the customer is in a good mood. That timing is why a surprisingly large share of buyers answers it: it takes three seconds and the order is already placed.

The classic is the channel question: where did you first see us? But the question that is pure gold for your creatives goes one layer deeper: what convinced you to order today? The answer comes in your customer's own words, and those words are exactly the raw material strong ads are made of.

Why is this the cheapest triangulation layer?

Since iOS 14, every platform misses part of the picture and fills the gaps with modeling. Meta reports what it may claim within an attribution window, not what your spend caused. Incrementality tests are the cleanest correction, but they cost time, budget and volume. A survey costs almost nothing, runs continuously and gives you fresh signal every week from the only source with no stake in the answer: your customer.

Be honest about the limitations. People do not remember every touchpoint, they name the last thing they saw, and some skip the question entirely. Survey data is direction, not bookkeeping. But that bias is stable. If the distribution of answers shows the same pattern for months and then suddenly shifts, the shift is real. That is exactly where the value sits: trends and breaks, not absolute percentages.

Which questions should you ask?

Keep it short and make the answers countable. This is the set we work with:

• Where did you first discover us? Fixed options per channel, plus an open field for everything else.
• What convinced you to order today? Open field, because here you want customer language, not a checklist.
• How long did you know us before your first order? This reveals how long your funnel really is.
• Where else did you see us after that? Optional, but useful for reading your channels as a whole.

Your customer is the only attribution model that tells you the truth for free.

How do you use the answers for channel decisions?

Every month, put the survey distribution next to the claimed revenue per channel. The patterns are often predictable: channels that work early in the funnel, like video on social, get structurally undervalued in click attribution, while branded search and retargeting get overvalued. The survey corrects that picture. If a channel gets named twice as often in the survey as the dashboard claims, you know where your budget headroom sits.

The question of how long someone knew your brand before their first purchase is at least as valuable. If the majority answers in weeks or months, your real funnel is far longer than your seven-day attribution window suggests. That changes how you judge top-of-funnel spend: those ads do work, just outside the window the platform is allowed to count in.

What does this mean for your creatives?

This is where the survey gets really interesting. The open answers to what convinced someone are a continuous source of angles and hooks. Customers literally quote back what moved them: a specific doubt that got resolved, a review that made the difference, a problem nobody else solved for them. We use this kind of customer language daily as input for new concepts, because a hook in your customer's words beats a hook from the meeting room almost every time.

Read the answers monthly and cluster them by theme. If a motif keeps coming back that no ad currently touches, you have found your next testing round. That way attribution data stops being a report after the fact and becomes fuel for the creatives that need to carry your next stage of growth.

Conclusion

A post-purchase survey does not replace your tracking, it keeps it honest. For almost no money, you get an independent data layer that corrects your channel mix and feeds your creatives with real customer language. Translating those insights into concepts and hooks that make strangers stop and buy is exactly the work we do every day. Curious what is hiding in your survey data? Book a call and we will gladly look at it together.