Perspective defines a shopper insight as the evidence-backed reason a shopper chose a product at the moment of purchase, and that definition is where most briefs go wrong. Not because it’s incorrect — Perspective’s framing is the standard one, and it’s a reasonable sentence to put on a slide. It goes wrong because a brief can pass that definition and still hand an agency nothing it can act on. The only insight worth the name tells you which of the three people in the aisle you’re talking to, and what that would make you build differently. “Dads use this at weekend barbecues” passes the Perspective test and still can’t brief a thing; “the partner buys it, and they’re checking it against the other brand on price in four seconds” names a different person and points at a mechanic — which means half of what gets filed as insight is really a consumer fact wearing a lanyard.
What is a shopper insight?
A shopper insight is an evidence-backed explanation of a purchase decision that names who made it — the shopper, not the consumer — and points at something you’d build differently as a result. The field defines it by the behaviour of the person actually shopping, not the end user. If it doesn’t change what you’d do, it’s a fact.
Why “why they buy” isn’t enough
Here’s the trap I see commissioned over and over: a shopper marketing manager signs off a research study before they’ve decided what the promotion is for. The study comes back rich, well-run, full of lifestyle and attitude and the emotional territory of the brand. And then it’s handed to an agency, and the agency reads it, and nothing in it says whether to build for trial or for basket size.
The reason is almost always the same. The study described the consumer — the person who uses the product, who is often not even in the store — and called it shopper research. It never separated the person holding the wallet from the person the product is for. So when you go to use it, it can’t point at a lever, because it was never watching the right person. This is the shopper versus consumer split, and it’s the single most expensive thing to get wrong before a brief.
There’s a third person the deck usually ignores entirely: the gatekeeper, the category manager who decided months ago whether the product would be on shelf at all. An insight that names which of the three people in the aisle you’re designing for points somewhere. An insight about the shopper changes the mechanic; an insight about the gatekeeper changes whether the promotion gets ranged in the first place — “the buyer won’t range a fourth SKU without a reason to believe it turns” is a finding you take to a ranging conversation, not to a creative team. A consumer occasion — “dads at barbecues” — changes a TV ad and leaves the shelf untouched.
The test: can you build from it?
The working test I’d give any insight is blunt. Read it, and ask what you would build differently if it were true. If the honest answer is nothing — if it’s interesting but it doesn’t move a single setting in the campaign — it’s a consumer fact, and you’ve paid research money for context, not direction.
An insight that names the shopper points at a lever you can actually pull. “They decide on price in seconds and don’t register the range extension” points at belief and friction at the shelf — a reason to make the claim legible fast, not to spend on a richer prize. “They’re loyal to the category but switch brands on promotion” points at trial, which is a different job from frequency, and a different mechanic entirely. The lever is the output a usable insight produces; lifestyle colour produces none.
Watch what the test does to a live example. Promo Monitor tracks 249 live promotions in Australia, and the single prize-draw is the plurality mechanic at 87 of them — about 35 per cent — ahead of instant win on 60 and gift-with-purchase on 34, so a draw is the thing a genuine insight has to argue past, not the safe default it looks like. The Vinarchy BWS Jacobs Creek Australian Open draw advertises a single $50,000 major prize; a tennis draw is a Headliner mechanic, built for fame and reach. That’s the right call only if your shopper insight is about awareness. If the real insight is “BWS shoppers already know the brand and switch on price at the fixture,” a draw builds nothing — the usable insight would push you off the sports draw and onto something that reads at the shelf in the four seconds you’ve got.
The counter-view: gather enough and the strategy writes itself
The fair version of the opposing case is stronger than it first looks. Shopper insights are the evidence-backed explanations of why a shopper chose a product at the moment of purchase; gather enough of them, properly segmented, and the picture gets dense enough that direction stops being a guess. A good shopper study already separates the switchers from the loyalists, the planned trips from the impulse ones, the price-led decisions from the range-led ones. A practitioner holding that evidence can reasonably say the job is self-evident from the data — you don’t impose an objective, you read it off the segment that’s moving.
Here’s where it breaks. Hand the same deck to two planners. Both read the same ten true observations: the category is bought weekly, a third of buyers are new to the brand, promotion-driven switching is high, loyal buyers rarely trade up, basket size is small, decisions are fast, and so on. Planner A reads “a third are new, switching is high” and writes a Breaker brief — a trial mechanic to convert first-timers. Planner B reads “bought weekly, loyal buyers rarely trade up” and writes a Keeper brief — a loyalty mechanic to lift the regulars. Same evidence, opposite mechanics, both defensible. The deck didn’t rank its own findings, because true things don’t rank themselves. Something has to decide which job the promotion is doing, and that something is a judgement about commercial priority this quarter, not a finding in the data. It doesn’t help that the moment of choice is now split across channels — IBISWorld’s 2026 analysis cites Australia Post’s eCommerce report putting 9.8 million households shopping online in 2025, up from 8.2 million in 2019 — so “the shelf” is several shelves, and no amount of gathered evidence tells you which one your job lives on. You decide the job, then the insight you need sharpens into focus. Not the other way round.
So the definition on the slide isn’t wrong, it’s just not finished. A shopper insight explains a purchase decision — but the test of whether you actually have one is whether it names the person at the shelf and changes what you’d build. Once the job is decided and the mechanic is chosen, the entry collection and prize fulfilment and winner payments are the delivery half of the work.
Misfile a consumer fact as a shopper insight and you pay three times: for research that can’t brief anything, for a mechanic built for the wrong person, and for the ranging conversation with the category manager you never knew you needed to have.
If you’re sharpening the insight before you write the brief, we’d welcome that conversation.
