Why AI-Native Companies Need More People, Not Fewer

The claim

The prevailing story about AI and jobs is subtraction: adopt AI, need fewer people, keep the same output, take the margin. I've spent the last decade inside a large retail business, moving from UX into digital transformation and AI workflows, and more recently built a live AI platform of my own. Across all of it, I've watched this play out enough times to think the story is backwards. The companies getting hurt by AI aren't the ones with too many people. They're the ones who used AI to make their existing structure smaller instead of building a different structure.

Three companies, one sector

Imagine three insurance companies, same sector, same basic business model, whether that's life or auto doesn't matter.

Company A is large: over a thousand employees, a traditional structure that predates AI by decades. Company B is roughly the same size. Company C is an AI-native startup, about ten people, built from day one around AI rather than retrofitted for it.

All three look at AI. Company A adopts it. Company B adopts it. Company C has no choice: AI-native workflows are the reason it exists.

Where Company A goes wrong

Company A's move is the obvious one: adopt AI, cut headcount, protect margin. The problem isn't the adoption. It's that the structure underneath survives the cut untouched: the workflows, the approval chains, the way the company creates value for a customer are all still the ones built for a thousand-person traditional company. Company A just runs that same structure with fewer people in it.

That structure has real, fixed costs. Salaries alone can put Company A's overhead north of £30 million a year. Cutting staff shrinks that number, but it doesn't restructure it. Company A still needs tens of millions in revenue just to stay partially profitable, because the shape of the company never changed, only its size.

What Company C actually proves

Company C isn't ten people doing the work of a thousand through effort. It's ten people inside a structure built for AI from the start: different workflows, different overhead, a different relationship between headcount and output. Company C's overhead might sit around £700,000. It only needs to clear a million pounds to be profitable.

That's not a headcount story. It's a structural one. The mistake is looking at Company C and concluding "you need fewer people to win." The actual lesson is what an AI-native cost structure looks like, and that structure is available to a thousand-person company too, if it's willing to rebuild around it instead of just cutting from it.

Company B's move

Company B also adopts AI. But instead of firing five hundred people to chase Company C's efficiency, it does the harder thing: it trains its thousand employees and restructures the company around the same AI-native principles Company C was built on.

Here's the arithmetic that matters. If ten AI-native people can produce what Company C produces, that means every ten people, restructured properly, are dramatically more productive than they used to be, not less necessary. Applying that multiple to Company B's existing thousand employees doesn't shrink the company. It makes it the equivalent of a ten-thousand-person company, without adding a single headcount.

What this means for Company A

Company A, having downsized without restructuring, isn't now lean. It's now competing against a company an order of magnitude more productive per person, with a workforce it never had to build from scratch. Company A cut its way toward Company C's size and missed that Company C's size was never the point. Company C's structure was.

The restructuring question

None of this means Company C-style startups lose. It means companies like Company A have a choice that isn't "adopt AI" versus "don't": it's restructure or shrink. Training and restructuring an existing workforce around AI-native workflows is a harder, slower project than a layoff. It's also the one that compounds. The companies that treat AI as a reason to need fewer trained people are solving the wrong problem. The ones that treat it as a reason to need more people trained to work inside a new structure are the ones about to outcompete both the giant that downsized and the startup that started small.