A project starts six weeks late because the senior architect is finishing two other things. A junior fills in for the first month. The seller does not notice. The PM does not notice, because capacity says everyone is available next month. Delivery notices, complains in standup, and gets told it will equalise.
It does not equalise. By the time the first invoice goes out, three weeks of senior effort have been spent at junior cost-but-junior-output. Margin missed by four points. The post-mortem blames "scope creep". The slip happened before scope was even touched.
This is one of the most common margin leaks in a services firm and it has nothing to do with estimation accuracy.
Utilisation tells you who's busy, not whether the right people are
Utilisation is the metric every services org reaches for when asked "are we resourced well?". It is also the wrong question. A resource pool at 90% utilisation can be 90% wrong-grade-on-wrong-deal and the dashboard will glow green.
What utilisation hides:
- Grade fidelity. Whether the people working the deal match the grades the deal was priced with.
- Skill match. Whether the people available have done this kind of work before.
- Ramp time. Whether the people on it now were on it from the start, or showed up three weeks late.
- Continuity. Whether the same people stay through the first delivery cycle.
A senior architect at 95% utilisation across three deals, none of which they were sold into at the rate they are now being charged out at, is a margin disaster wearing a green badge.
Resourcing velocity, in four parts
Resourcing velocity is the rate at which a sold deal becomes a staffed deal in the shape it was sold in. It has four components, all measurable at deal start.
1. Time to first allocation. Days between deal close and the first named resource on the project. Most firms do not track this. The gap can run anywhere from a week to two months.
2. Grade fidelity. Percentage of effort being delivered by the grade the deal was priced with, in the first four weeks. Junior-substitution is the silent killer here. If a deal was priced at 60% senior and the first month runs 30% senior, the grade fidelity is 50% and the margin is already gone.
3. Skill match. Whether the assigned people have shipped this kind of work before. A "senior" who is senior in another stack is a junior on this one. Hard to measure cleanly, visible in any honest practice review.
4. Continuity. Whether the same people stay on the deal through the first delivery cycle. Hand-offs in week three cost more than the equivalent ramp-up at the start, because by then the shape of the work is half-discovered and lives in someone's head.
Together those four are the leading indicator no one tracks.
What slow velocity actually costs
A worked example. A six-month project priced at $480k, sold with a margin target of 30%.
- Priced effort: 1,800 hours at a blended rate of $267/hr.
- Priced mix: 60% senior at $300/hr cost-loaded, 40% mid at $200/hr.
Now the deal starts. The senior architect is unavailable for the first three weeks. A second senior is pulled from another deal mid-project, four weeks in. For the first month, the actual mix runs 25% senior, 75% mid. Output drops, because mid-grade staff are figuring out architecture decisions a senior would close in a meeting.
Three things happen, none of them on a dashboard:
- Effort overrun. Mid-grade work takes 25% longer for the same output. The first month consumes about 380 hours of effort against a planned 300.
- Rework. Architecture decisions made by mid-grade staff get revisited when the senior arrives. Ten to fifteen percent of month-one work is reopened.
- Schedule drift. The project is now two-and-a-half weeks behind, which the PM eats into the buffer that was meant to absorb scope creep.
Margin lands at 21% on a deal sold at 30%. The seller hits quota. The PM gets blamed for the overrun. The senior architect gets told to be more available next time. None of those is the actual cause.
The actual cause: the deal was sold with no visibility into resourcing velocity. The pipeline did not see the senior architect was already over-committed. Sales did not see what they were selling against the same person another seller had already sold. The first the system noticed was the green-badge utilisation report at month-end.
This is a quote-time problem, not a delivery-time problem
By the time a project is in delivery, resourcing velocity is already locked in. The fix has to land at the quote.
Three things make resourcing velocity visible early.
Pipeline-aware capacity, not utilisation reports. A weekly view that combines committed work, forecast work (won deals not yet started), and exposed work (in-flight pipeline weighted by probability) at the role and grade level. Not just headcount. If three deals are pricing senior architect time for the same eight weeks, the system should show it before the third one is sold.
Capacity-aware selling. Sellers seeing role-month density before they commit to a start date. A practice lead seeing which roles are about to spike. Both are quote-time interventions, not delivery-time ones. Covered in more depth in capacity-aware selling.
Deal gating on resourcing confidence. A signed deal is not a green light if the resourcing-velocity calculation says month one will be 30% senior against a 60% commitment. The decision should be: defer the start, renegotiate the mix, or accept the margin hit explicitly, not by accident.
Two of those are workflow changes. The third is a tooling change.
Where this gets visible
The capacity, drivers, and revenue layers in forecasts put the role-month picture in front of sellers and practice leads at quote time, not at month-end. The drivers view shows which deals are creating the role-month peaks, which is the question utilisation reports never answer. Combined with schedule on the deal (best-fit duration calculated against the role mix you actually priced) the resourcing-velocity question becomes a quote-time conversation, not a post-mortem.
Wrapping up
Utilisation tells you the bench was billable. Resourcing velocity tells you whether the deal you sold became the deal you delivered. Most margin leaks in a services firm happen in the first month, before scope ever changes, and they happen because the resourcing question was answered with a number that was not asking it.
The metric you do not track is the one you cannot fix.
