Three different jobs wearing the same title
"Quantum engineer" covers work that has almost nothing in common beyond the machine it eventually runs on. The people are drawn from different disciplines, trained differently, and rarely move between these tracks once established. Deciding which one you are hiring is not a detail to settle later — it determines where you advertise, what you can pay, and whether anyone qualified reads past the first paragraph.
Quantum Software Engineer
Compilers, SDKs, simulators, control software and the tooling that turns a circuit into something a device can run.
Strong conventional software engineering. The most fillable of the three, and the one most often over-specified.
Quantum Hardware Engineer
Building and characterising the device itself — qubits, control electronics, the cryogenic and photonic stack around them.
Experimental physics, hands-on lab work, usually a PhD. Bound to a building.
Quantum Algorithm Engineer
Whether a problem admits a quantum advantage at all, and what the circuit for it looks like.
Theory depth and publications. The smallest pool, and the hardest to assess without a specialist.
Those three are the usual decision, but the field subdivides further and the narrower titles have their own, much smaller pools: Quantum Error Correction, Quantum Control Engineer, Cryogenics Engineer and Photonics Engineer among them. If your problem is specifically one of those, name it. The audience is smaller but it is the right audience, and a generic title will not reach them.
The mistake that keeps a role open for a year
The most common self-inflicted wound in quantum hiring is a specification that asks for both halves of the field in one person: a physics PhD with publications, who is also a production software engineer with a decade of distributed systems behind them. That person exists in roughly the numbers you would guess, and they are not reading job adverts.
The fix is almost always to split the requirement rather than raise the salary. A quantum software role that needs someone to build a reliable SDK, a testing harness or a compiler pass is a software engineering job with a domain to learn — and strong software engineers learn the domain routinely, given a physicist to ask. Insisting on the PhD for that role removes most of the people who could do it well, in exchange for depth the work never calls on.
If you cannot name the specific task that requires the doctorate, it is a preference rather than a requirement — and in a pool this small, preferences are expensive.
Who you are actually competing with
In most fields the competing offer is another company. Here it is often a university. A postdoc considering your role is comparing it against continuing an academic track they have spent a decade on, and the things they are weighing are not all financial — whether the work is real, whether they can publish it, and whether it closes off a return to research.
UK quantum work also concentrates geographically to an unusual degree. The clusters around Oxford, Cambridge, Bristol, London and Glasgow, along with the national programme at Harwell, account for the large majority of it. That cuts both ways: a role anchored to one of them can recruit from a genuine local pool, and a role anywhere else is competing at a distance for people whose spouses, labs and collaborators are already somewhere specific.
What the market is paying
Ranges in this field are wider than in most, because the same title covers a doctoral specialist and a software engineer who joined from a general engineering background. Treat the band as the spread across a role, not as a target for one person, and price against the shape you actually settled on above.
What you can offer that money cannot
Because so many candidates come from research, several of the strongest levers here cost little and are rarely offered explicitly. Stating them in the advert also signals that you have hired from this pool before, which is itself reassuring to someone leaving academia for the first time.
Say what they may publish
- A stated publication policy beats a vague assurance at interview.
- Conference attendance is how this field keeps its network — budget for it and say so.
- Silence is read as "no", and for a researcher that is a career decision.
Be specific about the machine
- Which modality you are building on, and at what scale.
- Whether they touch real hardware or work against simulators.
- What is already working, rather than what is planned.
Before you publish
Most of what makes a quantum advert work is subtraction — removing the requirements that were aspirational, and replacing the field-wide description with the specific problem this person would own.
Check the advert against this
If the person you want is coming straight out of a university — which, in this field, is more likely than not — the process itself needs adjusting as well. Hiring quantum talent out of academia covers what those candidates weigh and how to assess someone whose only record is a thesis.
Where to look first
Before writing anything, it is worth seeing how the rest of the field is pitching comparable roles — the seniority language, the ranges, and which organisations are hiring against you right now.