The number most founders look away from
Ask a consulting or tech-services founder what their bench is costing them this month and most will give you a rough figure with a wince. Ask them what their bench is supposed to cost, as a deliberate, planned number tied to their forward pipeline, and the room goes quiet.
Bench cost is almost universally treated as a symptom: the thing you fix by winning more work or letting people go. That framing is wrong, and it is quietly destroying margin in firms between £5m and £30m in revenue. The firms running at 32% to 35% EBITDA are not carrying less bench. They are managing it with more precision than the firms running at 22%.
What the P&L is actually telling you
In a consulting or tech-services firm, the cost structure is relatively simple. You have people, you have overhead, and you have the spread between what clients pay and what delivery costs. Utilisation is the dial that connects people cost to revenue output.
At 75% billable utilisation across a delivery team, every pound of salary generates roughly £1.33 in chargeable time value before any margin. At 65%, that drops materially. The firm looks roughly the same from the outside. The P&L tells a different story.
But here is what most founders miss: bench cost is not simply a function of utilisation. It is a function of timing. A consultant sitting on the bench for three weeks between a project rolloff and a new start is a very different cost profile from a consultant who has been unplaced for eleven weeks because the pipeline did not convert as expected. The P&L records both the same way. The business generated them completely differently.
Bench cost is not a utilisation problem. It is a pipeline visibility problem wearing utilisation's clothes.
The two types of bench cost that firms conflate
Separate them and the management actions become obvious.
Structural bench is the cost you carry deliberately: new hires in onboarding, staff between client assignments with a confirmed next engagement, people in internal capability-building that has a defined commercial payoff. This bench is an investment. It has a return. Firms that grow past £15m without losing margin are usually very intentional about what goes in this bucket.
Speculative bench is the cost you accumulate because the pipeline did not land the way you expected. A proposal that was 80% likely and then went quiet. A contract extension that did not come through. A client who paused scope. This bench is a signal, not a budget line. It is telling you that your pipeline visibility is weaker than your resourcing decisions assumed.
The problem is that most P&Ls do not distinguish between the two. They show you one bench number, you feel bad about it, you try to bill the people faster or cut them, and you repeat the cycle next quarter. Nothing structurally changes.
Why the 32% firm runs differently
The firms we work with that consistently hold 30% to 35% EBITDA do three things that the 22% firms do not.
First, they track pipeline conversion by stage with enough granularity to make resourcing decisions four to six weeks ahead, not four to six days ahead. When a £150k project is at confirmed-scope-not-yet-signed, the 35% firm has already provisionally allocated a consultant. When it slips two weeks, that consultant does not land on the bench unplanned. There is a buffer role, an internal project, a training commitment that absorbs the time and maintains productive cost.
Second, they have a cost-per-bench-week figure that sits alongside their average day rate in the management pack. Not buried in the P&L, visible on the same page as pipeline and utilisation. When bench cost per week rises, the conversation is immediate. When it is buried three levels deep in the accounts, it surfaces at the quarterly review when the damage is already done.
Third, they price the bench risk into their commercial model. Fixed-price engagements in particular carry roll-off risk: the project ends, the next one is not yet sold. The 35% firm builds a margin buffer into fixed-price work that accounts for expected average bench time between engagements. The 22% firm prices to win and absorbs the bench cost as a surprise every time.
The practical fix is not about headcount decisions
The instinctive response to rising bench cost is to stop hiring or to let people go. Both are sometimes right. But they are downstream decisions. The upstream fix is pipeline discipline and resourcing rhythm.
Run a weekly resourcing call that puts named consultants against named opportunities at each stage of the pipeline, not just confirmed work. Assign an owner to every gap in the forward schedule. Make speculative bench visible as a category, not hidden in overall utilisation numbers. When a consultant is unplaced for more than two weeks with no confirmed next engagement in the pipeline, that is a commercial conversation, not an HR one.
On the commercial side, review your day-rate assumptions for the margin you actually need. If your blended day rate is £700 and your average bench time between engagements is three weeks per consultant per year, that bench cost needs to be covered by the billing weeks, not written off as variance. The maths is straightforward. Most firms have not done it.
The maths is straightforward. Most firms have not done it.
What this means for EBITDA and valuation
A ten-person delivery team with an average fully-loaded cost of £75k per head and three weeks of average annual bench per person is carrying roughly £43k of speculative bench cost per year. At 20 people, that is £86k. At 40 people, it is £172k. In a firm doing £8m revenue, £172k is two EBITDA points. In a transaction at a 7x multiple, that is £1.2m of enterprise value sitting in a line that most management teams treat as background noise.
The firms that close that gap before a sale are not the ones that cut headcount. They are the ones that connect their pipeline process to their resourcing process and stop treating bench as an unfortunate weather event.
Call this the Invisible Bench Drain: the steady, untracked bleed between pipeline slippage and resourcing decisions that sits just below the threshold where anyone calls a meeting about it. In the £5m to £30m band, it is one of the most consistent suppressors of EBITDA we see. It is also one of the most fixable.
If you want to understand where your firm sits across the seven value drivers that include operating model discipline, the Equity Snapshot takes three minutes and benchmarks you against top-quartile firms in your category. Take the diagnostic and see where the points are coming from.