There is a decision being made in a great many companies this year by people who do not know they are making it. It is being taken one workflow at a time, in the ordinary course of removing cost from a process, and it will not appear on any agenda under its real name. The real name is: we have decided what our senior bench looks like in six years.
I want to be careful about what I am not arguing. I am not arguing that firms should pay people to do work a machine does better. That is sentimental and no board should accept it. Nor am I arguing that the entry-level work being automated was good work — most of it was tedious, much of it was done badly at two in the morning, and the people doing it did not enjoy it.
The reps were the product, and the output was the by-product
The argument is narrower and I think harder to dismiss. That work was doing two jobs. It produced an output, which is the job everyone measured, and it produced a person, which is the job nobody did. When you remove it you get the first saving immediately and you incur the second cost invisibly, on a delay long enough that whoever authorised the change will have moved on before the consequence is legible.
I have watched this exact structure play out once already, and so has everyone reading this. Companies removed a management layer, booked the saving, and spent the following years discovering that the layer had been where people learned to run things at small scale before running them at large scale. The rebuilding of the middle is now a line item at firms that cut it, and it costs more to rebuild than it saved to remove, because you are buying back a capability rather than a headcount.
The version of this that worries me most is not the large firm. Large firms will notice eventually, and will pay to fix it. The version that worries me is the professional services firm of forty people where the founder does the hard work, three seniors do the rest, and the juniors have been replaced by a subscription. That firm has no bench, no succession, and no mechanism to build either — which is the problem founders assume they solved by hiring an operator, arriving early and from below.
What I would ask of any executive signing off on one of these changes is a single question, and it is not whether the automation works. It is: where does the person who does this job in ten years get their reps? If there is an answer — a structured programme, deliberate rotation, review work done at volume, anything — then proceed, because the answer is what makes it a decision rather than a drift. If there is no answer, the honest position is that the firm has chosen to buy its senior people from someone else's pipeline.
That has been a viable strategy for individual firms and it cannot be one for all of them at once. If everyone automates the training ground and plans to hire experienced people, the experienced people stop being produced, and the shortage that follows will be described in exactly the terms we always use — a talent shortage, a market failure, an unforeseeable squeeze. It will be a training shortage, foreseen, and chosen.
Topics opinionaiworkforcemanagement
