A denominator cannot choose a capability
Utilisation is attractive because it turns a complicated operating question into a percentage. In a services business, it can help reveal unassigned capacity and prompt better scheduling. It becomes less useful when every hour outside customer billing is treated as the same kind of failure. An engineer waiting for an assignment, an engineer helping a colleague recover a difficult implementation and an engineer learning a skill needed by future customers may all appear on the same side of that ratio. Management still has to decide which activities it wants to fund.
At Avelrune Digital Services, we considered a programme involving 155 engineers and 12,400 paid training hours. Using the agreed internal cost rate of EUR50 an hour, the programme represented EUR620,000 of allocated labour cost. The hours were not free because the people were already employed. Some could otherwise have supported billable work, although we did not assume that every available hour would certainly have been sold. Making that distinction visible helped us discuss both the opportunity given up and the limitations of assigning it a simple revenue value.
Delivery leaders believed that developing existing colleagues would serve the next phase of customer work better than relying entirely on outside recruitment. Some commercial managers preferred immediate assignments followed by hiring when demand became certain. Both positions had merit. Training too early can produce skills that sit unused; hiring too late can leave a signed commitment without a credible delivery team. Finance could not resolve that uncertainty by choosing the highest short-term utilisation figure. We needed an account of the customer demand, the capability gap and the timetable on which each option could respond.
The first proposal used course completion as its principal measure of success. I asked for a second question: what work would trained colleagues be able to undertake, under whose supervision, and when might that work become available? This was not an attempt to make every learning activity prove immediate commercial payback. It was a way to distinguish a defined capability investment from a general aspiration to become more skilled. The programme could include broader development aims, but those aims needed their own explanation rather than being hidden inside an unsupported revenue forecast.
Paying for time without promising a return
We approved the paid learning time and made the capacity consequence visible in the operating plan. The relevant managers could not be expected to deliver the same immediate billable schedule while releasing the same people for training. Some sessions were moved to reduce clashes with customer milestones, but the plan did not pretend that careful scheduling eliminated the cost. Managers identified the service commitments that had to remain covered and the supervision that newly trained colleagues would need. That preparation was as important as the purchase or design of the learning material.
I did not take responsibility for declaring an engineer technically proficient. Delivery leadership owned that assessment. Finance asked whether the planned hours, staffing cover and subsequent assignments were consistent with the investment case. We also separated attendance from capability and capability from commercial use. A colleague could complete a course without being ready to work independently, or become capable before an appropriate assignment existed. Treating those stages as interchangeable would have made the programme appear more successful than the available evidence supported. It would also have made it harder to diagnose a disappointing result fairly.
Within six months, 118 of the 155 engineers had used the relevant skills on customer work. Thirty-seven had not yet done so. We kept both numbers in the review. The observation showed deployment within a defined period; it did not establish that all 118 were equally proficient, that the remaining thirty-seven had learnt nothing, or that training alone generated a particular amount of revenue. Some assignments would have occurred with a different staffing approach. Customer demand, project timing and supervision all influenced the outcome. The review therefore retained those qualifications alongside the numerical result.
There was pressure to convert the deployment figure into a financial return. I resisted a single payback claim because the counterfactual was weak. We had not run an experiment comparing otherwise identical teams with and without the programme. Instead, we examined specific assignments, avoided recruitment needs where these could be supported, and the continuing gaps in the delivery plan. That evidence informed the next funding decision without pretending to measure everything the programme had caused. A CFO can require discipline while accepting that some strategically relevant choices will remain only partly measurable.
Follow the deployment, not only the course
The thirty-seven colleagues without a relevant assignment became a useful starting point for management action. In some cases, the anticipated customer work had moved. In others, the next task required experience that a course could not provide on its own. We discussed supervised assignments and the availability of senior colleagues rather than simply scheduling more training. The financial consequence included time from people who were already in demand. An investment in junior or adjacent capability can require additional capacity from the very specialists whose scarcity prompted the programme in the first place.
This is where a narrow utilisation discussion can become counterproductive. A senior engineer helping another colleague may reduce personal billable hours while improving the team's future ability to deliver. That does not make every mentoring request worthwhile, nor does it excuse poor scheduling. It means the manager should explain the expected benefit and the resources required at team level. Finance should make those commitments comparable with alternatives, including hiring, contracting, narrowing the service offer or accepting a slower growth plan. The ratio remains useful evidence, but it cannot choose among those operating strategies by itself.
I would improve the next programme by agreeing assignment opportunities earlier and recording the assumptions that would cause us to change its scale. A broad growth ambition is not enough. We need to know which customer work is sufficiently plausible, which skills are transferable if that work changes and which managers can provide supervision. I would also make the review period explicit before the programme begins. Without that agreement, a team can be criticised for slow deployment against a deadline it never accepted, or claim success by extending the observation period indefinitely.
For directors, the question is whether management can explain a deliberate reduction in immediate capacity as a considered use of resources. The answer should include the paid hours, the cost convention, the commercial opportunities forgone, the evidence of subsequent deployment and the remaining uncertainty. In our case, EUR620,000 and 12,400 hours belonged in the account as clearly as the 118 colleagues who used their new skills. That balance helped us retain the development ambition without describing it as costless. People investment deserves the same honest treatment of alternatives and limitations as any other consequential business decision.