Pack redesign: when doing nothing is the win

Most pack changes get approved in a room where everyone has stared at the current pack for years, and the shopper it is aimed at has never looked at it for more than a moment. That asymmetry is the problem.

Front of pack labelling proposals have Indian food brands opening the artwork file right now, and once it is open the temptation is to change everything else too. Sorting a forced change from a preferred one takes an hour, and almost nobody spends it.

Forced, or preferred

Write the trigger down in one sentence, then sort it. A change is forced when something outside your preference makes the current pack unusable: a regulatory requirement lands on a label element, a substrate is discontinued, a manufacturing move changes the available format. Where the trigger is regulatory, get the scope of what actually has to change confirmed by your own legal or regulatory advisor, because the requirement is usually narrower than the redesign it ends up justifying.

Everything else is a preference, and preferences are legitimate. They simply have to compete on evidence. “The pack looks dated” is a preference. So is “the new marketing head wants their own pack”, and “a competitor relaunched” is a preference wearing a strategy jacket.

The sorting question: if this change does not happen, what goes wrong, and in which number would we see it? If the honest answer is that the brand feels tired to the people who work on it, you have found a real feeling and not a business case.

The refresh your team can see and a shopper cannot

Your team reviews the pack at A3 on a wall, or full screen in a deck, for weeks. A shopper meets it as a thumbnail in a scrolling grid, or in peripheral vision while looking for something else. Almost nothing that reads at A3 reads at thumbnail.

What survives that reduction is short: the dominant colour block, the silhouette of the pack, and the single largest element on the face. What does not survive is longer: a new typeface, revised kerning, a shift in illustration style, a redrawn mark that keeps its proportions, a repositioned descriptor. Those are the changes design reviews spend most time on and shoppers register least.

Test that before you argue about it. Render the proposed pack at the pixel dimensions it will occupy on the platforms you sell on, put it beside the current one and look at both on a phone. If you cannot tell them apart at that size, you are about to pay for a change no shopper will notice. Quick commerce grids are the constraint worth designing against, not the wall.

Recognition is the asset you are spending

A repeat buyer does not read your pack. They navigate to a memorised shorthand, usually one colour and one shape, fast enough that they never consciously check. That shorthand took years and a lot of media money to install.

Change it and the repeat buyer has to find you again. Some do it instantly, some take a few trips, and a few pick up the thing next to you. The cost is real and close to impossible to attribute afterwards: it shows up as a soft dip in repeat rate that gets explained away by an algorithm change or a competitor promotion. If you track repeat buyer cohorts, set the baseline before the change rather than after someone asks what happened.

The uncomfortable part: the more recognisable your pack is, the more a redesign costs you, and the more likely your own team is bored of it.

Change one element, not the system

List your distinctive assets in order of how strongly they are recognised: usually colour, then pack shape, then the brand mark, then a device or character, then typography, then layout, then photography. Change from the bottom of that list upward, and only as far as the trigger genuinely requires.

This matters most when the trigger is forced. A required element on the front face is a layout problem, not a licence to redraw the system. Usually it can be accommodated inside the existing architecture by giving up something less recognised, and the pack still looks like itself the week after, which is also faster when the changeover has to be sequenced against stock you already hold. The same logic applies to the recipe: if the trigger pushes you toward reformulation, ask which single attribute has to move, not how much else can be improved while the file is open.

Test whether it solves a measured problem

A pack change can be tested before it is tooled, and it is one of the few brand decisions where that is cheap. Run the proposed render as creative in a live ad unit against the current pack and read the difference in click through and conversion. Show both to people who buy the category and have never seen your review deck, through a properly built survey rather than a group chat. Ask existing buyers to find their usual pack in a timed grid of competitors and see whether the new one slows them down.

None of that is expensive next to tooling. All of it gets skipped when the change is a preference, because a preference does not want a result that might contradict it.

Doing nothing deserves someone to argue it

The cost of changing is knowable. Build it with what a pack change actually costs and you will have a real number per SKU. The benefit of changing is almost always asserted rather than measured. A decision where one side has a number and the other has adjectives should not be close. It usually goes to the adjectives, because nobody in the room has the job of defending the pack that already works.

Give someone that job out loud. Their brief is narrow: what the current pack is still doing well, what recognition is at risk, and what else the change budget would buy. If the change survives that, it is probably real.

Deferring is not refusing. Tie the decision to the next natural tooling or reorder event so the option stays open at a fraction of the cost. If the answer when you arrive is that a SKU is not worth re-tooling, that is a portfolio question rather than a pack one, and it belongs with rationalisation. When the change does go ahead, sequence the telling with telling the channel before you change the pack and plan the overlap with running two pack versions on shelf.