Verification is the least interesting slide in any campaign report and the one everything else rests on. A trial you counted and a trial you estimated look identical on a chart. They are not the same claim, and only one of them survives being questioned.
What we actually do
Proof of purchase is matched to the customer and to the content they post, and checked by our team.
Three things have to line up: the person who was matched to the brief, the purchase they made, and the content that came back. Only when all three match do we count the trial, license the content and pay for it as delivered work.
That is the whole mechanism. It is deliberately boring, and the boringness is the feature.
Why the purchase is a purchase
The obvious question is why customers buy the product at all when a brand could simply send it to them.
Three reasons, and they are commercial rather than philosophical.
A basket is a decision. Somebody who chose your product, in a shop, alongside forty alternatives, has done the thing the campaign is actually trying to cause. Somebody who opened a courier bag has not.
It produces a receipt. A physical artefact tying a real person to a real transaction at a real retailer on a real date. Nothing about a gifted product produces that.
It changes how people talk. A customer who spent their own money talks about a product differently from someone who was given it, and shoppers can hear the difference immediately.
The customer is refunded what they spent and paid for the content, so they are not out of pocket. The purchase still happened.
What verification makes possible
Every other number in the report depends on this one.
- The trial count becomes a count rather than an estimate, which is the difference between a number a category manager accepts and one they discount.
- The retailer read becomes possible. If the brief needs a specific banner, receipts prove it happened there.
- The content is defensible. A customer testimonial from somebody with no proven purchase is a liability under section 12A of the Fair Trading Act, which prohibits representations made without reasonable grounds at the time.
- The evidence file exists. The dated original statement, the proof of use, the incentive record, the final edit and the substantiation, held together for every asset.
The failures it catches
Most of what verification catches is mundane rather than dishonest, and worth naming so the numbers are not mistaken for a fraud rate.
Receipts for the wrong variant. Purchases made outside the campaign window. Two products bought and submitted as separate entries. Somebody who bought it, genuinely loved it, and posted about it three weeks after the window closed.
Each of those is a legitimate trial for the customer and an invalid one for the report. They are excluded anyway, because a count that makes exceptions is not a count.
What it costs
Time, mostly ours. Checking receipts against content against participants does not scale the way a media buy does, and it is one of the real costs of running a verified sampling programme rather than a gifting programme.
The question to ask any sampling proposal
Not how many people will try it. Ask how many trials will be verified, how, and what happens to the ones that cannot be.
If the answer is that trials are estimated from product distributed, the number in the final report is a division sum rather than a measurement. That may be fine for a launch update. It will not hold up in a trading meeting, and it should not.
How the rest of the process runs is on product sampling, and the moderation step is covered in what we reject, and why.