Why asking if they “like” your new packaging is a terrible idea

If you’re gonna quant test your new packaging, it’s kinda important to measure things that actually correlate with in-market success.

Designalytics has been doing packaging research for a looooong time. Recently, they did some research with IRI to see which test metrics predict in-market results for 52 pack redesigns across a range of CPG categories.

They compared sales for the 6 months after the redesign with the same 6 month period from the prior year. Then they checked how well those sales results correlated with the different metrics from the pack test.

What they DIDN’T predict in-market success too good was whether people “liked” the new design. (It had a 46% correlation.)

Surprisingly (for me at least), find time wasn’t correlated with sales results either.

So what metrics DID matter?

First, purchase preference over the stuff they currently buy. That had a whopping 96% correlation.

And second, whether the design communicated the key purchase drivers of the category. That had an 88% correlation.

Which, after all, makes sense: if you’re selling ice cream, does the pack look yummy and creamy? And does it make you want the ice cream more than what you currently get?

Designalytics says they’re still checking whether standout is a good predictor, but that’s a bit harder to measure.

So when you do your next pack design, ask the questions that matter!

Some Lessons:

🔸 Don't rely on “liking. Just because people like a package doesn't mean they'll buy it. 

🔸Measure preference over current packs. If your design beats what people already buy, you're onto something.

🔸Make the category benefits obvious. If it's ice cream, it should look craveable.