Headhunting for Niche Roles: Finding People Who Do Not Show Up in Search

You run the search. Twelve results. Two look plausible, one is at a competitor you cannot touch, and the other has been in the same job for nine years. The hiring manager asks how it is going.

Before concluding the market is empty, it is worth checking which of three quite different problems you actually have — because they have different fixes and only one of them is genuine scarcity.

Three problems that all get called "niche"

Problem	What it looks like	Fix
Genuinely small population	A few hundred people worldwide with this exact background	Widen geography, or accept training someone in
Vocabulary mismatch	They exist in numbers but do not use your words	Take your vocabulary from them, not the job description
Hiding in adjacent industries	The skill exists elsewhere under a different name	Map which industries produce the capability

In practice the second and third are far more common than the first. Genuine scarcity exists, but it is rarer than the phrase "impossible role" suggests.

Technique 1: take your vocabulary from a real person

The job description was written by someone inside your company. The candidate's profile was written by them, about their own career, using the words their industry uses. These two documents frequently do not overlap.

So do not start from the job description. Start from one person.

1. Find one person who is obviously right. A past placement, someone the hiring manager admires, a competitor's team member. They do not need to be reachable — you are using them as a specimen.
2. Read how they describe their work. Not their title. The verbs and nouns in their actual descriptions.
3. Write down every term that surprised you. The ones that were not in your job description.
4. Rebuild the search from those terms.

This takes about twenty minutes and routinely doubles or triples the pool, because you stop searching for how your company talks and start searching for how the market talks.

What this looks like in practice

A brief asks for a "Technical Program Manager, Infrastructure". Searching that title returns a thin list dominated by very large companies. Read three people who genuinely do the job and you find they describe themselves as delivery leads, engineering programme managers, or simply senior engineers who ran a migration. The capability is common; the title is not.

Technique 2: search for what they built, not what they are called

Titles are assigned by employers and vary wildly. The work is the stable thing.

Instead of the title, search for the artefacts: the system, the migration, the certification, the regulation, the specific tool. Someone who has done the work will mention the work, whatever their title says.

Instead of searching	Search for
"Payments Product Manager"	The scheme, settlement, reconciliation, chargebacks
"Regulatory Affairs Manager"	The specific regulation or submission type
"Platform Engineer"	The specific migration or the tooling involved
"Clinical Data Manager"	The trial phase, the standard, the system

This also filters usefully by depth. People who genuinely did the work describe the specifics; people who were nearby describe the category.

Technique 3: map the adjacent industries

Ask the hiring manager one question: where else does this capability get built? They usually know, and it is usually not written down anywhere.

The pattern is that a skill developed under pressure in one industry exists, unlabelled, in several others. High-volume transaction handling exists in payments, in ticketing, in gaming and in logistics. Rigorous documentation practice exists in pharma, aviation and defence. Working under real-time constraints exists in trading, telecoms and broadcast.

Each adjacent industry is a fresh pool where you are not competing with everyone else hiring for your job title. The trade is that these candidates need a reason to believe the move makes sense, which is an outreach problem rather than a sourcing one.

Technique 4: work backwards from the pattern

This is the highest-yield technique and the least used.

Take two or three people who would obviously be right. Ignore their titles entirely and look at the shape of their careers:

• What kind of company did they start at?
• What was the pivotal move — where did they learn the thing that matters?
• How long do they stay?
• What did they do just before the role that made them good at this?

You are looking for a repeatable pattern. Something like: started in a large regulated company, moved to a small one at around 50 people, stayed through significant growth.

That pattern is a much better search than any title, because it describes how the capability gets built rather than what it ended up being called. It is also the kind of criterion no filter can express — you cannot tick a box for "moved from large to small and stayed through growth", which is precisely why the people matching it are not being contacted by everyone else. Our piece on what boolean cannot express covers why.

Technique 5: go where the work is published

For specialised fields, general profile databases are often the worst source, because the strongest people are visible through their work rather than their profile.

• Conference speaker lists. Anyone who spoke at a specialist event has been vouched for by a programme committee.
• Professional bodies and association directories. Often public, often complete for regulated professions.
• Published work. Papers, patents, open-source contributions, standards committees.
• Trade press. The people quoted in industry publications are the recognised practitioners.
• Alumni of the obvious companies. If three companies are known for this capability, everyone who left them is a candidate.

Five names from a conference programme can be worth more than five hundred from a broad search, because the filtering has already been done by people who understand the field.

When the pool really is small

Sometimes it genuinely is thirty people. That changes the arithmetic of outreach completely, and most recruiters carry over habits that only make sense at volume.

With three thousand possible candidates, a message that gets ignored costs almost nothing — you send more. With thirty, each person is effectively a single attempt. A generic message that gets ignored has not cost you nothing; it has removed a name from a list that cannot be replenished.

So for small pools:

• Slow down. Twenty minutes per message is proportionate when there are thirty people in the world.
• Reference the specific work. Not their title, not a recent post — the thing they built that made you contact them.
• Be honest about the constraints. Specialists can evaluate the role properly and will discover the awkward parts anyway. Naming them earns more credibility than hiding them.
• Accept a no gracefully. In a field of thirty people, they talk to each other. Today's no is a referral source, and sometimes next year's yes.

The structure of a message that earns a reply from someone not looking for a job is covered in sourcing passive candidates — the difference between specificity and personalisation matters more here than anywhere else, because you cannot compensate with volume.

The question worth asking the hiring manager

Before any of the above: which of these requirements would you drop for the right person?

Most niche briefs contain one genuine requirement and three inherited preferences that nobody has questioned. The distinction is rarely obvious from the job description, and asking directly usually reframes the search in a way no sourcing technique can.

A useful follow-up: if someone had everything except this, how long would it take them to learn it? If the answer is a few months, it was a preference. Splitting requirements from preferences is what makes the difference between listing four hard constraints and fourteen — and the process in building a shortlist from criteria depends on getting that split right.

Where tooling helps and where it does not

Techniques 1, 3 and 5 are judgement work. No tool replaces asking the hiring manager where else the capability gets built, or reading a conference programme.

Technique 4 is different. Once you can describe the pattern — the career shape that produces the capability — you need to find everyone matching it, and that is exactly the search no filter can run. This is where Hiris fits: you describe the pattern in the same words you would use to a colleague, and get back the people whose history matches it, with the evidence for each criterion attached so you can check the reasoning rather than take it on trust.