Packaging design testing before launch checks whether a new pack will be seen on shelf, understood at a glance, and remembered and felt the right way – while there is still time to change it. For UK CPG brands, the most efficient approach is to screen every concept and variant with AI creative testing in minutes, then send a proven shortlist to consumer research. Dragonfly AI decodes Attention, Memory and Emotion for packs on their own and in real UK shelf scenes.
Most UK consumer packaged goods brands test their flagship pack and send everything else to print on judgement. Packaging design testing before launch changes that. It gives brand teams evidence on whether a new pack will stand out, communicate clearly and convert – before artwork is locked and the print run is booked.
This guide covers what pre-launch packaging testing should measure, when to do it in your development timeline, how the main packaging concept testing methods compare, and how to combine fast AI testing with traditional UK market research.
Pre-launch packaging design testing is the process of evaluating pack concepts and final artwork with evidence before they reach shelf. It answers three practical questions: will shoppers notice the pack, will they understand what it is and why to choose it, and will they remember the brand next time they shop the category.
It sits alongside wider pre-launch product testing rather than replacing it. Product testing checks what's inside – taste, performance, price. Packaging testing checks what shoppers see first, often in a split second as they scan a crowded fixture. A strong product in a pack that disappears on shelf rarely gets the chance to be tried.
Launching in UK grocery is hard and getting harder. Kantar reports that almost two-thirds of new launches are dead or declining by the end of their second year. Own-label now accounts for more than half of UK grocery volumes for the first time, according to Circana data reported by The Grocer, so most branded packs sit right next to a retailer equivalent designed to look just as credible.
At the same time, consumer packaged goods brands are changing packs more often for reasons they didn't choose. Under the UK's packaging Extended Producer Responsibility (EPR) scheme, producers now pay fees based on how much packaging they place on the market and how recyclable it is – a strong push to switch materials, lightweight formats and redesign. Every one of those changes alters the pack face, and can quietly cost it standout.
The upside of getting it right is real. NielsenIQ BASES research shows an optimised pack design can lift forecast volume by up to 5.5%, at no extra cost per unit once printed. But the same research finds that nine in ten redesigns fail to make a meaningful sales impact. Testing is what separates the two.
A useful pre-launch test measures specific outcomes, not general opinion. "Do you like it?" is the weakest question you can ask about packaging. Focus on these five instead:
Calbee's experience shows why each one needs measuring separately. When it temporarily moved some packs to black and white, Dragonfly AI's analysis showed Attention held at 100%, but Emotion fell from 84% to 20%. A pack can stop the eye and still fail to make anyone feel anything.
The biggest mistake in packaging design testing is doing it once, late, when the only options left are "launch" or "delay". Build it into each stage of development instead.
|
Stage |
Key question |
Best method |
|
1. Concept and design routes |
Which routes are worth developing? |
AI creative testing on every route |
|
2. Design refinement |
Which elements are helping or hurting? |
Element-level diagnostics on logo, claim, hero image and variant cues |
|
3. Shelf context |
Will it stand out against UK competitors? |
Shelf scene testing with the real competitor set, including own-label |
|
4. Shortlist validation |
Do shoppers understand, trust and want it? |
Consumer research – surveys, qualitative work, purchase intent |
|
5. Pre-print sign-off |
Has anything slipped since approval? |
Packaging design validation – rescore final artwork against the agreed benchmark |
Early stages need speed and volume, because you're choosing between many options. Later stages need depth, because you're making a commitment. The final check matters more than most teams realise: legal copy, nutrition information and recycling symbols are often added late, and each one squeezes the design. If a mandatory change makes the pack less visible or less clear, it shouldn't be approved without a plan to recover what's been lost.
Each method answers a different question, so the right choice depends on the stage you're at.
|
Method |
What it tells you |
Turnaround |
Best for |
|
Focus groups and qualitative research |
Why people react as they do – meaning, language, associations |
Weeks |
Exploring positioning and early ideas |
|
Online surveys |
Appeal, comprehension and purchase intent across a representative sample |
Days to weeks |
Validating a shortlist |
|
Eye tracking and virtual shelf studies |
Where real shoppers look, on screen or in store |
Weeks |
High-stakes relaunches of a lead SKU |
|
AI creative testing |
Attention, Memory and Emotion at concept, element and shelf level |
Minutes |
Screening every route and variant, and refining designs |
Cost shapes how these get used. Eye-tracking and shelf studies typically cost around £22,000 to £37,000 per design and take weeks, which is why most brands reserve them for the lead SKU. Traditional research remains essential for questions only people can answer. The problem is using expensive tools on questions that don't need them, which leaves too little budget for the ones that do.
The most efficient approach for UK CPG brands is a funnel. Screen every concept, variant and format with AI testing early, when a result takes minutes. Improve the designs that make the cut using element-level diagnostics. Then send two or three finalists to consumer research, rather than twelve. Your UK market research budget goes much further on a shortlist that has already proved it can be seen.
This also makes it realistic to test the long tail – the flavour extensions, multipacks, seasonal editions, shelf-ready packaging and export variants that usually launch untested, yet together shape what shoppers think of the brand.
Dragonfly AI decodes how real people will see, feel and remember a pack before it's printed. Brand teams use it in three ways:
Want to see how your next launch performs before it's printed? Book a demo here.