There is a particular kind of pride I keep noticing at the moment. A founder announces that they spent a Saturday with an LLM and came out the other side with a name, a logo, a colour palette, a tone of voice document and a live website. Eight hours from nothing to something, for the price of a monthly subscription. The comments underneath are full of congratulation, and I understand why. It looks like resourcefulness. It looks like the sort of momentum that gets a business moving.
I should be clear that I am not standing outside the tent throwing stones. We use AI at Ótimo every day and it has genuinely changed how we work. The argument I want to make is not about the tool, it is about where people are pointing it, and about the quiet assumption underneath all of this: that a brand is a set of assets you can acquire, rather than a position you have to earn.
Buying has never been a rational act
Ask anyone why they chose the supplier they chose and they will give you a reason that sounds like logic. Better price, better spec, better availability. Watch what they actually do and something else is going on entirely. They chose the company whose emails felt like they were written by a human being. They chose the one whose website suggested somebody had thought carefully about their problem before they arrived. They chose the one that behaved, in a hundred small ways, like an organisation with a point of view.
You know this from your own life. You have paid more for coffee because of the room it was served in. You have stayed with a supplier who was not the cheapest because leaving felt like a loss. Preference is emotional first and justified afterwards, and that has been true for as long as anyone has been selling anything to anyone else.
Which means the job of an identity is not decoration. The job is to signal, fast and consistently, what a company believes and how it is likely to behave. Nobody has ever felt anything in particular about a logo that could have belonged to nine hundred other companies.
An average cannot differentiate you
It is worth understanding what these models are doing when you ask them for a brand. They have been trained on an enormous quantity of work that already exists, and they generate the most statistically probable next thing. That is the mechanism, and it is a remarkable one, but it has a consequence that matters enormously here. What you get back is a considered summary of the middle of the market.
So the logo arrives looking like a consolidation of every other logo in your category, because in a real sense that is exactly what it is. The website arrives as a competent average of every other website in your category: the same hero statement, the same three benefit columns, the same testimonial band, the same closing call to action with a gradient behind it. The tone of voice arrives sounding warm and approachable and expert, which is what everybody else's tone of voice document also says.
None of it is bad, and that is the trap. It is all perfectly acceptable, which is why it is so easy to ship and so hard to argue with. But the middle of the market is the one place a challenger cannot afford to stand, because standing there means asking a customer to notice you while presenting them with the visual and verbal equivalent of everyone they have already ignored.
The saving that is not a saving
The cost argument is the one I hear most, and I understand it, particularly for a young business watching every euro. But it only works if you count one side of the ledger.
Say a properly considered identity and site costs a meaningful sum and takes real weeks. Set against that the customer who visited, felt nothing, and went to the competitor whose site made them feel something. Set against it the investor who formed a view of your seriousness in eleven seconds. Set against it the senior hire who looked you up, saw a company that appeared to have been assembled rather than built, and never replied. Set against it the discount you will end up offering because you have given the market no other reason to choose you.
An identity generated in an afternoon is a cost reduction, and it is worth being honest that this is all it is. It reduces a line item. It does not generate revenue, it does not compound, and it does not build anything you own. Distinctiveness is the only marketing asset that gets more valuable the longer you hold it, and an average has nothing to compound.
The sea of same
Zoom out from your own business for a moment and look at what is happening across every category at once. Thousands of companies are drawing from the same models, feeding them broadly the same prompts, and receiving broadly the same answers. The result is a market that is converging on a single look and a single voice at a speed nobody has seen before.
For most of the businesses in that convergence, this is a slow disaster. For the small number who refuse to join it, this is the best commercial opportunity in years. When everyone else is drifting towards the centre, the cost of being genuinely different collapses. You no longer need to outspend anyone. You only need to be recognisably yourself, consistently, while your category dissolves into a beige middle distance.
Where AI genuinely earns its place
I promised this was not a rant against the tool, so let me be specific about where it belongs.
It is excellent at research, at reading the market faster than a team of people could, at pattern-finding across categories, at pressure-testing a position by arguing against it, at drafting, at the operational and administrative work that quietly eats a founder's week. We use it constantly for exactly these things and the work is better for it.
The distinction I would offer is this. Use AI on the parts of your business where sameness costs you nothing, and protect the parts where sameness costs you everything. Efficiency where you compete on execution. Craft where you compete on meaning.
The window is open now
Here is the part I would ask you to sit with. Your competitors are, at this moment, in conversation with a language model, pleased with themselves for having built a brand before lunch. They will post about it. They will feel, for a few months, as though they have got ahead.
You should be doing the opposite. Slower, more considered, more specific to you, more difficult to copy. Decide what you actually believe before you decide what you look like. Behave like it everywhere, including in the places customers do not expect anyone to be paying attention. Hire somebody who has spent a career learning where the difference lies, and give them the time to find it.
That work does not pay back this quarter. It pays back for years, and it pays back most in exactly the conditions we are heading into, where everyone else has quietly agreed to look the same.
Craft now. Gains later.
Ian Howarth
September, 10, 2026
Every AI company has landed on the same logo. Go through thirty of their websites in an afternoon and you'll lose track of which one you're on, because it's the same soft geometric mark or gradient orb every time, and the word intelligence is always up near the top doing very little.
