Testing a Package Before It Ships

A new package has two jobs before anyone tastes what is inside. It has to be found on a crowded shelf in a few seconds, and it has to say something a shopper wants and believes. A packaging and claims test measures both before the design goes to the printer, while changing it is still cheap.

The question it answers

Most redesigns are judged in a conference room, on a large screen, one package at a time. Shoppers never see a package that way. They see it at arm's length, next to a dozen competitors, for a second or two. The test asks three questions under conditions closer to the shelf: can people find it, what does the eye go to first, and which claims earn a place on the front panel.

How we run it

Our standard format is online. Each respondent sees one design only, so the designs never compete with each other in the same head. A typical study uses about 150 category buyers per design. They start with a timed task on a simulated shelf of a dozen products, with the test package rotated across positions. We record whether they find it, how long it takes, and what they pick by mistake. Then they see the package on its own and click the first thing that catches their eye, which builds a heat map of attention by area: logo, flavor name, claim and imagery. Last, everyone rates a list of candidate claims for appeal and believability, and a reach analysis, often called TURF, finds the small set of claims that together appeal to the most people.

What you get

A side-by-side scorecard for every design, the shelf and attention results behind it, the claims ranked two ways, and a recommendation that names the trade-offs instead of hiding them. Each gap in the scorecard is marked as significant or directional, so nobody argues over a difference that could be noise.

Online or in person

Most package design tests run online: hundreds of shoppers, results in days, and attention measured by where people click first. When the question is about the shelf itself, such as a new layout, a fixture, a display or where a brand sits in the aisle, we also run the test in person. We build full-size shelf sets in our facility, and shoppers wear lightweight eye-tracking glasses that record where they actually look, at real distance and in real light. The output is a gaze heat map and a spotlight map of the real shelf, plus scores for every package on it: how many shoppers looked, how fast and for how long. Then they shop the set, so you see what that attention turned into.

In-person studies are smaller, usually 20 to 30 shoppers per layout, so we read them as directional and often pair them with an online test to size the effect.

Where it goes wrong

The most common mistake is putting the three highest-scoring claims on the pack. Top claims often appeal to the same people, so three of them reach barely more shoppers than one. The second is treating appeal as the whole story: a claim people like but do not believe can cost more than it adds. The third is reading small gaps as wins. With 150 people per design, a three point difference in purchase intent is noise, and the report should say so. Finally, a test cannot speak for a shelf it did not model. If the set at your key retailer looks different, we build that shelf.

The first sample below is the online version: a fictional sparkling water brand testing its current can against two redesigns, with 450 respondents split evenly across the three designs. Click the tabs to move from the shelf task to the heat maps, the claims and the recommendation. You can even try the shelf task yourself.