The next generation of package design screening combines artificial and human intelligence

The future is supervised machine learning, trained with behavioral consumer data. When a brand launches a new product or wants to redesign a current one, they’ll likely have a lot of different options to assess and test. Until now, validation tools that provide a simple yes or no answer are the standard.
Blog

But putting new designs straight into validation means you’re more likely to fail than you are to succeed. That’s because this process occurs too late in the design process.

When we looked at our full database, we found that when brands do their design screening early, 65% of new designs ultimately outperformed the current design, compared to just 40% with late stage screening - in other words, validation.

Early-stage design screening drastically increases your chances of in-market success. That was the insight that drove our latest innovation - the AI Pack Screener.

Because it comes in earlier in the process, the AI Pack Screener is perfect for when you need to redesign but you aren’t sure which direction you want to go

Previously, early stage also meant lots of exploration - which could be very costly to test.

Just imagine how expensive and time-consuming validating 20 design concepts would be. We can narrow down dozens of ideas to the few with the most potential whilst being quicker and more cost-effective than validation. You have to screen beforehand to increase success in the later stages.

But can you trust an artificial intelligence system with your new product or packaging launch?

Of course, no AI is perfect. There’s a lot of scepticism and confusion around artificial intelligence. For that reason alone, it’s fair to question it; AI has its limitations.

We know that AI is only as good as the data it’s been trained with. If an algorithm is built with poor training data, its output will also be poor.

Algorithms are also designed to provide generalizable outcomes. That means they’re more likely to choose a solution that benefits the most people. But this means AI can be blind to specific context for particular brands.

No AI will ever be able to overcome this; you simply can’t train an AI for every possible outcome, and change that for each brand study.

And while algorithms can provide us with data, they can’t tell us how to act on that information in a meaningful way. There’s only so far machines can take us.

So how can you overcome AI’s limitations?

The AI Pack Screener is the only tool of its kind to combine artificial intelligence with human intelligence. We call this The Human Advantage™️.

How it works:

Design changes are recorded and benchmarked against our extensive normative database of over 30,000 packaging designs. That training data is rooted in real behavioral consumer data, too, so you know it’s trustworthy.

Then, we take it one step further. We involve our panel of experts to do what a machine will never be able to - provide context, educated opinions and informed recommendations.

Our panel of experts can tell you how this data works in the context of your brand, and guide you to act on it in a meaningful way. They’re able to go beyond the numbers and provide comprehensive advice that is completely tailored to your brand and its objectives.

Only real behavior can predict real-life outcomes. And only real people can provide holistic perspectives on index numbers and probability scores.

 

Author

Blog psot author avatar

Olivier BLANCHET - CEO at PRS IN VIVO,