What the Jaguar rebrand says about scaling with AI

Whether it was good is not something we can settle here, and sales effects are not ours to claim. What makes it a useful example is something narrower and more borrowable: it is a rare, fully public case of a brand changing direction loudly, at full volume, across every channel at once. Most rebrands happen slowly enough that you never see the seams.

And the seams are the interesting part, because they are exactly what happens to a product brand when AI arrives in its content operation.

The confusion underneath

Here is the mistake we see in our own market, phrased as a sentence somebody actually said to us: "Now that we can make everything, we need to decide what to make."

That is the wrong order, and it is the order most teams end up in. The capability arrives first, because it arrives as a tool and tools can be bought on a Tuesday. The direction is supposed to arrive first, because direction is what makes volume worth anything.

A rebrand makes this visible in a way that a slow drift never does. Everything changes at once, everyone sees it at once, and the question of whether the direction was decided well becomes unavoidable.

AI in a content operation produces the same visibility, just without the announcement. When your team goes from forty images a month to four hundred, every unresolved question about your brand gets answered four hundred times, by whoever happened to be at the keyboard.

What AI scales, and what it does not

It scales output. More images, more variants, more markets, more channels, at a cost that no longer forces you to choose.

It scales consistency, if you have rules. The same setup, applied the two hundredth time exactly as it was applied the first. This is genuinely new, and it is the best thing about it.

It scales your ambiguity too. If your rules are vague, you now get vague output faster and in larger quantities. The drift that used to take three years takes one quarter.

It does not scale direction. Nothing automates the decision about what your brand should look like, for whom, and why. That decision is made by people, in a meeting, with disagreement in the room, and it does not get faster or cheaper.

So the honest summary is: AI removes the production constraint and leaves the judgement constraint exactly where it was. If judgement was your bottleneck before, you have not solved anything. You have just made the queue longer.

Who guards the direction when everyone can make images

This is the organisational question, and it is the one most teams have not answered.

Before: making an image required a photographer, a studio, a budget line and three weeks. Those frictions were annoying, and they also functioned as a review process. Nobody produced an off-brand image by accident, because producing any image at all required somebody to approve a cost.

Now: a product manager in another country can generate a usable image in ten minutes, for a local campaign, with the best intentions. Multiply by the number of people in your organisation who have both a deadline and a login.

Three roles to name explicitly, before this becomes a problem rather than after:

The owner of the rules. One person who decides what an exception is. Every exception either joins the reference set or is refused. Not a committee; committees produce exceptions by accumulation.

The approver of output. Someone who sees what leaves the building, or a defined sample plus everything in a risk category. At volume, "everyone keeps an eye on it" means nobody does.

The owner of supply. Who makes sure partners, dealers and markets receive the approved set rather than finding something themselves. This one is almost never named, and it is where most brand damage actually happens, on pages you do not control.

What to fix first

If your organisation is about to add volume, in this order:

1. Write the fixed core. Five or six decisions that make your product recognisable. One page, with a reference image per rule. We described how in rules a production process can follow.

2. Write the negative rules. What must never appear. These catch the plausible mistakes, which are the ones that get published.

3. Name the three roles above. By name, not by department.

4. Define what complete means per product. Then produce complete sets rather than more of the same image.

5. Check outside, not inside. Your own site is the one place where everything was done on purpose.

Step five keeps finding the expensive problem. We wrote about the rest of it in what breaks first when you scale.

Our own position, plainly

We make AI-produced visuals for brands, so we have an obvious interest in you producing more of them. That is exactly why we would rather say this part out loud: volume without direction is not a win, it is a faster way to look inconsistent.

It is also why we keep a person from our team in the loop on everything we deliver, and why we start from a Brand DNA rather than from a prompt. Not because automation is bad, but because the thing AI cannot do is decide what your brand should be. You can see what that produces in our use cases.

In closing

A loud rebrand and a quiet AI rollout create the same test: can your organisation say what it wants to look like, specifically enough that two hundred people and one production process reach the same answer?

If yes, volume is an advantage. If no, volume is just a louder version of the uncertainty you already had.

Want to work out what your fixed core is before you scale? Try LensLab free with five credits, or have a look at our use cases first.