The great equaliser.

In all the IT revolutions before, the right answer was: get bigger.

The shape of corporate competition since the late 1990s favoured scale at almost every layer. Bigger budgets bought better software, bigger teams bought specialisation, bigger procurement bought longer roadmaps. The companies that grew fastest were the ones that compounded these advantages, and the gap they opened over smaller competitors was, for a long time, real.

Inside those companies, however, the people doing the actual digital and innovation work were rarely operating from a position of strength. They had the budget, but speed and internal agreement were often missing. The friction that kept small competitors at a distance was, on the inside, the very thing that slowed them down.

That friction has now become the most expensive asset they own.

What changed is the floor.

The work that used to require a small department can now be trusted to a single person with the right toolset. Things like research synthesis, data engineering, copy and design, analysis, monitoring, it has all changed. But it's not just visible in the work that people do, it's also in the tools and software companies use. Especially at smaller scale, most of it can quite easily be built.

This is not a claim about future AGI. It's a claim about what's ready today. The infrastructure exists, the interfaces are usable, and the pricing for the kind of work most teams do is so low it's almost trivial.

The implication is uncomfortable for the multinationals and freeing for everyone else. The advantage of having a large team is being eroded by the fact that the small team can now do most of what the large team used to, faster and quite simply better.

Where the day actually goes.

If you take a single piece of work, like a small clinical reporting workflow or a regulatory monitoring brief, and time-budget how it gets done in two environments, the picture is consistent. The multinational team spends most of its hours coordinating; the small team spends most of its hours doing the work.

This is not a moral observation. The multinational has coordination overhead for reasons that are individually defensible, like compliance, audit trail and alignment across functions. The cost is real, and so is the benefit. But once those costs sum up, the actual production time, the share of the week in which someone is making the thing better, turns out to be remarkably small.

This should be the real question.

The blunt "AI will replace X jobs" framing is, frankly, nonsense. It's looking at the wrong question entirely. If you're asking yourself the question "How can AI make my current work cheaper, more efficient and faster", you're asking the wrong question. The real opportunity AI offers lies in these questions: how much more can we do, how much better can we be, what value can we add for our customers, if we learn how to apply AI to our daily work?

These are not speculative questions. They are observed, in the field, in teams where one person now does the work that would until recently have required several and where software and application are being built by people who are specialists in their field, but not in coding. What matters for a small team is that the people they already have are now capable of much more, given the time and the right shape of help.

The advantage is open now. And visibly closing.

The honest version of this argument has to include the closing line: the multinationals are not asleep, and they are catching up. The friction inside large organisations is being slowly worn down by procurement reform, by internal AI platforms, and by a generation of leaders who watched what happened to the previous generation and don't want it on their watch.

That gives small teams a window that is wide right now, narrower in eighteen months, and substantially closed inside three years. The work is to compound: a team that builds two new workflows this quarter and three next quarter will be in a very different place by 2028 than one that runs a single pilot.

We work with the teams who take this seriously.

If the argument lands, if you read this and recognise your own team in the small-team column of the figures above, then the next question is straightforward. Do you treat the next 18 months as a normal year, or as the year where you decisively pull ahead?

The work we do is the second answer. We help teams in this window build the habits, workflows and tools that compound: the kinds of small daily improvements that, after eighteen months, look like a different company.

How we work →    What you could build for yourselves →