Every consultancy and agency I have worked in ran on the same arithmetic. A senior person wins the work. A pyramid of people underneath does it. The margin lives in the ratio between the two.
David Maister, who wrote the book on running professional service firms, called it leverage: the number of junior people each senior one can keep busy. It is the quiet engine of professional services. The client buys a senior name and a senior judgement, and most of the hours on the invoice come from people learning the job on the client’s time. Nobody really pretends otherwise. It works because the work at the bottom of the pyramid, the research, the analysis, the first drafts, the decks, the documentation, needs a lot of hands.
Small teams. Agents doing the work that doesn’t matter. People owning the calls that do.
What the agents take
That is exactly the work agents now do well. Not perfectly, but quickly, cheaply, and better every few months without anyone asking them to. Research that took a team a fortnight takes an afternoon. A first draft takes minutes. A working prototype can replace the requirements document it would once have taken three people to write.
I have seen this film before, at a smaller scale. Early in my career I spent late nights drawing redlines for developers. A plugin took the job, then a tool removed the task entirely, and the value we had created by doing it went with it. That was one task in one discipline. This time it is the whole bottom of the pyramid.
When that layer can be done by agents for a fraction of the cost, the arithmetic stops working. A firm built on leverage is selling hours that no longer need to exist, and its clients will notice long before its partners want them to.
What a client is actually buying
It helps to pull apart what a consultancy sells. There is intelligence: the research, the analysis, knowing what is going on. There is judgement: deciding what matters, what to do and what to leave out. And there is accountability: someone who will stand behind the recommendation, carry it through the organisation and own what happens next.
The pyramid bundled all three and charged by the hour. Intelligence is the part being commoditised fastest; you can buy a remarkable amount of it for the price of a subscription. Judgement and accountability are not being commoditised at all. They are also the parts that were always done at the top.
So the value hasn’t gone anywhere. It has just stopped needing the scaffolding underneath it.
Leverage moves, it doesn’t vanish
The title is slightly unfair, and I should say so. What ends is leverage through junior hours. Leverage itself moves.
It moves into systems: agent workflows that do the volume, a shared record of the work that every agent reads before it writes, design systems and brand guidelines written for agents to follow, prototypes that become specifications. A small senior team can now direct work that used to need a floor of people. And unlike a floor of people, the systems get better with every engagement instead of walking out of the door when someone leaves.
That changes what a firm is. The old one was a people business with some tools. The new one is a small team of people with judgement, running on a platform they keep improving. Hiring stops being about delivery capacity and starts being about orchestration, trust and taste.
What changes for the client
The first thing is that you stop paying for learning on your time. The people in the room are the people doing the work, and the people doing the work are the people who decide.
The second is price. Hours stop being an honest measure when an agent does a day’s work in minutes, so time-based pricing starts to look like what it always partly was: a way of charging for inefficiency. Outcomes are the honest unit. That is harder to sell and harder to scope, which is precisely why it is worth doing well.
The third is speed. When the prototype is the specification, the gap between deciding and seeing shrinks from quarters to weeks. Decisions get better because people can argue about something real rather than something described.
The hard part
There is a cost, and it would be dishonest to skip it. The pyramid was also how people learned. Juniors became seniors by doing exactly the work agents now do: the research, the drafts, the endless revisions. If the bottom of the pyramid disappears, where does the judgement at the top come from in ten years?
I don’t think the answer is to keep people doing busywork for the sake of their development. It is to treat apprenticeship as an explicit job rather than a by-product of utilisation. People learn by reviewing, steering and owning decisions alongside someone experienced, while the agents do the volume. That takes deliberate effort, and most firms built on leverage won’t make it, because the old model never required them to.
The other cost is accountability. When agents do the work, it becomes tempting to let them make the calls too. That is the version of this future I want no part of. Agents do the work; someone accountable makes the decision. The moment nobody owns a decision, you have rebuilt the worst of the pyramid without any of the people.
The bet
The firms that last won’t be the ones with the most people. They will be the ones with the best judgement, the best systems and the clearest accountability, and they will be a lot smaller than the ones they replace.
Small teams. Agents doing the work that doesn’t matter. People owning the calls that do. That is the bet I left to test, and the one AltCo is built on.