Forward capacity against pipeline, and why utilization is the wrong signal.
Headcount forecasting is the largest single margin lever available to a professional services firm. Every other improvement is worth a few points on one engagement. Getting hiring and ramp-down timing right against your pipeline operates on your entire cost base.
The reason it dominates is arithmetic. An engagement running at 35 percent instead of 40 percent costs you five points of one project's revenue. Five people on the bench for six weeks costs you five full salaries against nothing, and it lands in overhead where no individual engagement shows it. This is how a firm ends up with every project looking profitable and the business making no money.
The number most firms watch is current utilization, and it describes the past. It tells you how well you matched capacity to work in a period that has already closed. You cannot act on it, because the decision it would have informed was made months ago.
Hiring runs on a long lead. Recruiting, notice periods, and ramp-up together mean the person you decide to hire today is productive in something like three to five months. A signal that arrives when utilization is already high arrives too late to be a hiring signal. It is a description of a problem you are now inside.
Forward capacity against contracted and likely work, six to twelve weeks out and further where you can.
That requires three inputs that normally live in three places:
Synthelio brings all three together, which is the whole argument for connecting your CRM rather than treating it as a sales tool that lives elsewhere.
This is the detail that decides whether a forecast is useful or decorative.
A deal worth 600,000 closing in November is not November revenue and it is not November capacity. If it kicks off in January and runs six months, it is roughly 100,000 a month from January through June, and it is six months of people you need to have.
A pipeline that only carries a value and a close date tells you nothing about when you need bodies. That is why the CRM field mapping asks for an expected kick-off date and a number of months of delivery. Without them, forecasted revenue collapses into the month a deal closes, which is the one month it is guaranteed not to be earned in.
Three shapes, and each calls for a different response.
Capacity exceeds committed and likely work, and the gap grows. A bench is forming. The lead time on the response is short, so this is the one you can still act on: push sales at a specific capability, pull forward internal work, or slow hiring.
Committed work exceeds capacity within your hiring lead time. You are already late. The options left are subcontracting, hiring at a premium, or telling a client a date they will not like. Choose deliberately rather than defaulting to overtime, which is the expensive option that feels free.
Committed work exceeds capacity beyond your hiring lead time. This is the good outcome, and it is the entire point of forecasting forward. You have time to hire properly.
The failure here is optimism, and it is structural rather than personal.
Your CRM probabilities are a sales instrument. They are set by people whose job is to be confident, and they tend to overstate. Hiring against unweighted pipeline is how firms end up with a bench and a hiring freeze in the same quarter.
Two habits fix most of it. Compare forecast to outcome for a few quarters and derive your own conversion rate rather than trusting stage probabilities. And separate the decisions by confidence: hire permanently against contracted work, and cover probable work with contractors or subcontractors until it is signed.
Firms treat hiring as a decision and ramp-down as an event that happens to them. It is the same forecast read in the other direction, and it is worth deciding in advance what you will do when an engagement ends with nothing behind it.
The choice is usually between carrying the cost, redeploying, or reducing. All three are legitimate. Making the choice six weeks out is a strategy; making it the week the engagement ends is damage control.
A person's employment FTE is currently stored as a single value rather than a dated history. If you change it, for example moving someone from full time to four days a week, historical utilization is recalculated as though the new figure always applied.
It does not affect forward planning, which is what this article is about. It does mean that a utilization figure you exported before such a change will not match the same figure recalculated afterward, so keep that in mind when comparing periods.
Tell us what you need. We will point you to the answer or write the article.
