A second screen in the same market reaches a cleared expert materially faster than the first, and the third faster again. The saving is real and almost never booked, because screening time sits inside a project fee rather than a line of its own — so nobody sees what starting from zero costs.
What does starting from zero cost?
Time, mostly, and it is invisible because it sits inside a project fee rather than on a line of its own. A practice can pay it repeatedly for years without a number ever appearing that would prompt a different decision.
It also costs optionality. A screen run under deadline selects for whoever is reachable, which is not the same as whoever is best on the call that follows.
The invisibility is what makes it durable. A cost on its own line gets questioned annually; a cost absorbed into delivery gets questioned never, because nobody is accountable for a number that does not exist. That is why the first step is measurement rather than change — a practice that starts recording screening hours against each project usually reaches the decision on its own within two quarters.
How much faster is a second screen?
Materially, and the effect compounds through the third and fourth. What accumulates is not a list of names but a picture of the market: which roles hold the knowledge, which companies to approach, which titles mislead.
That picture is the asset, and it is exactly the part that does not transfer when work is outsourced deal by deal.
The picture also degrades if it is not used. A market screened three times in a year and then left alone for two years is closer to a fresh start than the count of previous screens suggests, because the roles have moved and the companies have reorganized. What compounds is recent repetition rather than cumulative history.
Does a bench improve quality too?
Yes, and the quality effect is probably larger than the speed one. Repeated screening produces judgment about who is worth speaking to, which no list can encode.
It also improves the screen itself. Practices that have screened a market three times write better screens on the fourth.
Judgment of this kind is difficult to hand over, which is the practical constraint on treating a bench as an institutional asset. It lives with the people who ran the screens, and it leaves when they do unless something forces it into writing. Recording why a candidate was rejected, not only that they were, is the cheapest way to make it survive a departure.
Why does nobody book this cost?
Because it is absorbed. Screening happens inside delivery, and delivery is measured against the project fee rather than against what it cost to produce.
Making it visible is usually enough to change the decision, which is an argument for tracking it centrally even where nothing else changes.
There is a second reason nobody books it: the person best placed to measure it is the person whose delivery it would appear to slow. Asking a project lead to record screening time is asking them to document an overhead against their own engagement, and the incentive runs the wrong way. It works better as a desk-level measure than as a project-level one.
When is a bench worth building?
Once a practice expects three or more projects in the same market within a year. Below that, the maintenance cost of keeping it current exceeds the screening saving.
Above it, the compounding advantage is real and the practice is paying for it either way — once as a bench, or repeatedly as re-screening.
Below the threshold, the right answer is usually a partial bench rather than none. Keeping ten names current in the two segments a practice sells into most often costs very little and removes the worst version of a cold start, which is beginning a deadline-driven screen with no view of who is worth approaching.