Insights
The Hidden Workforce Skills Gap
A company I heard about recently spent nine months and a recruitment fee hiring a data scientist. The person they needed had been sitting in their operations team for four years. Nobody was at fault. She had joined as a logistics coordinator. Somewhere along the way she taught herself statistics, then Python, then built a demand model in her own time because the forecasting spreadsheet kept embarrassing everyone. Her manager knew she was good with numbers. That was the entire extent of what the organisation knew about her.
I keep returning to this story because it is not a story about talent. It is a story about vision. The organisation could see her output. It could not see her.
It is a story about vision
There is now enough evidence to say this is structural rather than anecdotal. In a March 2026 survey of 1,500 US managers and employees, 75% of managers said their team’s skills were fully utilised. Forty-nine percent of employees said their own skills were underused. Half said their company hires externally for skills that already exist internally. Two groups, one workplace, incompatible readings of the same reality.
A separate 2026 survey of HR leaders found the mechanism. Among those who described their workforce skills data as complete and accurate, three-quarters also estimated that fewer than 75% of their people’s skills were actually captured in any system. Roughly one in five update that data continuously. Asked what missed internal mobility costs them each year, 37% said two million dollars or more.
I want to be careful about the conclusion here, because the obvious one is wrong. The obvious conclusion is that organisations need better skills databases. Add a field. Run an inference model over performance data. Make the invisible legible.
That will help at the margin. But it misreads the problem.
An organisation does not fail to see a person’s capability because it lacks a form. It fails to see it because it only ever looks at one surface: the work it assigned. Everything an employee is, beyond the task they were given, falls outside the instrument. We built organisations to allocate execution, and allocation only requires knowing whether the assigned thing got done. For two hundred years that was enough, because a person’s assigned work and a person’s usable capability were roughly the same size.
They are no longer the same size. When execution becomes cheap, the assigned work shrinks and the residual — everything a person could think about, notice, question, connect — becomes the larger part of what they offer. Which means the organisation’s field of vision is now pointed at the smallest and least valuable portion of its own workforce.
The visibility gap is not a reporting failure that
This is what I find quietly serious about it. The visibility gap is not a reporting failure that got worse. It is a reporting failure that used to be affordable.
The logistics coordinator is not an unusual case. She is the ordinary case, observed. In most organisations she stays unobserved, does competent work, and eventually leaves for a company that asked her a different question in an interview. The nine months and the recruitment fee are not the real cost. The real cost is four years of a demand model that existed and was never used, because the person holding it had been categorised on her first day and never re-categorised since.
A record of the intelligence you already employ
I wonder whether the most valuable thing a leader can do this year has nothing to do with technology at all. It might be to sit with each person on their team and ask what they are capable of that nobody has ever asked them to do.
You will not get a database out of it. You will get something more useful: a record of the intelligence you already employ and have never once deployed.