Filling Niche Engineering Roles When Forty People in the Country Can Do the Job

On most roles, a better process finds better candidates. On a genuinely scarce role, every agency reaches the same small population and process changes nothing about supply. What it changes is who gets there first, who is credible enough to be answered, and whether the client will accept the trade-offs that the actual market requires. Those three things, not sourcing capability, decide the outcome.

Last reviewed August 2026. Delivery observations from Sortinghat customer base, stated as directional.

Establishing that the role is actually scarce

Most roles described as impossible are not. They are badly defined, badly priced, or over-constrained on location.

The test is a market map. Search on capability rather than title, remove one constraint at a time, and see what the population looks like at each step. Frequently the role opens up considerably when the location constraint softens or when one nice-to-have is dropped.

If the population is genuinely forty people, you now know something valuable and you can tell the client, which is a more useful first conversation than a shortlist of near-misses.

Why being answered matters more than reach

If every competing agency reaches the same forty people, then coverage is not a differentiator. Response is.

Three things move it. Somebody credible making the approach rather than a researcher with a template. A message that demonstrates understanding of what these people actually do. And a prior relationship, which is why your own database matters more here than any external source.

A person who spoke to your firm two years ago will read the message. A stranger with forty approaches in their inbox will not, which makes the depth of your own records the actual competitive asset on scarce roles.

The conversation the client has to have

On a scarce role, the client will have to move on something. The only question is which, and when they find out.

The four levers are compensation, location, the seniority band, and whether an adjacent skill set with a ramp-up period is acceptable. Every one of them is a real trade and clients rarely volunteer flexibility until presented with evidence.

The market map is that evidence. Telling a client that the combination they want exists in forty people, of whom eight are plausibly reachable, is a better use of week two than sending three candidates who almost fit.

Why the commercials have to change

Contingency terms assume that effort produces a reasonable probability of payment. On a forty-person population with several agencies competing, that probability is low enough that a rational firm should decline.

Retained or container terms fix the alignment. They also change client behaviour, which matters more than the fee: a client with money committed engages with the trade-off conversation instead of waiting to see whether somebody produces a miracle.

Firms that take these roles on contingency either abandon them quietly or work them badly, and both damage the relationship more than declining would have.

Adjacent skills, the answer that usually works

Most scarce-skill searches are eventually resolved by hiring somebody adjacent and accepting a ramp.

The useful preparation is to define, with the client, what adjacent means before the search stalls. Which fundamentals are non-negotiable, which domain knowledge can be learned, and how long the ramp is allowed to be.

That is a criteria conversation rather than a sourcing one, and doing it in week one rather than week eight is the difference between a filled role and an abandoned mandate. The mechanics are in how criteria get weighted.

Frequently asked questions

The map worth building before the shortlist

Before sourcing a scarce role, build the market map: how many people match on capability, how that number changes when you relax each constraint, and how many of them your firm has ever spoken to.

Present that in week two instead of three near-misses. It changes the client conversation from why have you not found anyone to which of these constraints should we move, which is the only conversation that ends with a placement.