Skawa
One of Poland’s oldest confectionery brands was choosing between four packaging designs. The team already had a favourite. We asked 600 consumers.
How often each design was seen as… · the “risky” design in ink, the “safe” favourite in grey
600 consumers · 3 phases · 1 week, against a standard quarter
The question. Which design should Skawa launch, and why? Inside the company, the classic design felt safe. No one at the table was the shopper the brand needed to win.
How we ran it.
Choice. A MaxDiff study: people picked the best and worst designs on six dimensions, from premium perception to ingredient communication. Two designs went through, two were eliminated.
Reasons. 40-minute AI-moderated interviews asked why: what people see in each pack, what the colours, type and heart motif mean to them, and what frustrates them about packs already on the shelf.
Validation. A representative sample rated one winning design each, close to a real shelf decision: price, purchase intent and the risk of cannibalizing Skawa’s other products.
What people told us. The “safe” favourite read as familiar but ordinary. The design tagged internally as “risky” was seen as modern twice as often (25% vs. 12%), as standing out twice as often (21.5% vs. 9%) and as premium nearly three times as often (15% vs. 5.5%). The heart shaped from biscuits read as someone having put their heart into it. The picture held across age groups.
What this shows. Choice data shows what people pick, and interviews show why. Collecting both from the same people, at scale and in a week, gives preference data with the reasoning attached. That is what reward models are missing today.
“You usually pick between scale and depth. This time we didn’t have to. 600 consumers gave us the scale to trust the answer. Reason8’s AI-moderated interviews gave us the depth to act on it. Together they change what research can do.”
Paweł Wikło
CEO, Skawa
Result. Skawa is launching the design it would not have chosen on its own.
Next case