DESIGN INTELLIGENCEJuly 22, 2026·Mary · DEPIX Design Intelligence

What a Redesign Must Not Touch

Every year the car industry ships a fresh crop of all-new models, and every year it measures the redesign the same way: by how new it looks. The 2026 model year suggests the metric is backwards. Some of the most thoroughly reworked cars are being rejected not for looking dated, but for looking like somebody else's car — and the pattern points to a design decision most redesigns get wrong at the very start.

Take the 2026 Jeep Cherokee. It is genuinely all-new: upright grille, squared body, chunky proportions borrowed from the Grand Cherokee. And it arrived to a very specific complaint — that its front end, in the words of one owners' forum, "screams GMC to me, not Jeep." Others mourned the boxy XJ generation as the "real" Cherokee. What is striking is that Jeep saw it coming. The company has said the single biggest risk in the whole program was redefining the Cherokee's visual identity, and it deliberately brought back the seven-slot grille to anchor the car to its lineage. It named the load-bearing cue, protected it — and a chunk of the audience still felt the identity slip.

That is the real lesson, and it isn't "redesigns are bad." It is that the hardest, highest-leverage decision in any redesign is not what to change. It is deciding what must not change — and being right about it.

Novelty is the easy part. Any studio can make a shape look new; the industry is full of new shapes. What is hard is the inventory taken before the sketching starts: which elements are the brand, and which are merely this generation's expression of it? A design language exists precisely to separate the two — the permanent grammar from the temporary styling. Get that inventory wrong and you can spend a fortune making something genuinely all-new while quietly deleting the one feature that made people recognise it and want it.

The 2026 field is full of the failure mode. Roundups of redesigns that "don't please everyone" almost always describe the same thing: a beloved trait removed, a face that no longer reads as the marque, a nameplate that lost the thread. It rarely comes down to the new design being ugly. It comes down to a broken promise — the customer bought a specific identity, and the redesign traded it for freshness.

How do you tell a load-bearing cue from a replaceable one? A useful test is subtraction: remove the element in a sketch and ask whether a loyal owner would still recognise the car as theirs — and still want it. The seven slots, a particular roofline, a signature light graphic — pull those and the whole identity wobbles. Trim, garnishes, this year's wheel pattern — nobody grieves them. The asymmetry is what raises the stakes: the people a redesign most easily loses are the loyalists, the hardest and most expensive audience to win back, and the ones who were keeping the nameplate alive in the first place.

This is a concept-phase problem, not a late-stage one. Brand continuity cannot be bolted on at the end, once the surfaces are frozen and the tooling committed. The decision about which cues are non-negotiable has to be made before anything is drawn, because it constrains everything that follows — the grille, the proportion, the silhouette a customer clocks in a mirror. Even the wave of new design languages arriving across the industry in 2026 will be judged less on how bold they are than on how well they decided what to carry over.

There is a trap on the other side, too. Over-protect every old cue and you get heritage pastiche — a costume of quotations rather than a living design. So the skill is sharper than "keep the old stuff." It is knowing which two or three cues actually carry the identity, and having the nerve to change everything else. A facelift or full redesign is really a curation problem: a short, ruthless list of what stays, and permission to reinvent the rest.

Every mature product has its seven-slot grille — a single element so bound up with identity that the audience treats it as the definition of the thing. Find yours before you redesign, not after the reviews land. Because the brands that redesign well are almost never the boldest ones. They are the ones who did the quiet, unglamorous concept-phase work of naming what they refused to touch — and then touched everything else with confidence.

Sources: