When a Sole Practitioner Can Actually Afford a First Hire

The email you're drafting begins "unfortunately I'm at capacity until March." It's the third one this quarter and the second one you actually wanted. Somewhere between the apology and the sign-off, the thought lands: I should just hire someone.

Hold that thought for a minute, because being too busy is not evidence that you can afford staff. It's evidence that you're too busy. Those are different problems with different solutions, and only one of them commits you to a fixed monthly liability that doesn't care whether planning goes to committee. The honest trigger for a first hire isn't workload. It's contracted, staged fee income far enough ahead, plus a cash buffer, plus a chargeable-hours sum you've actually done on paper. Here's how to build all three.

The Salary Is the Deposit, Not the Price

You've probably compared a Part 1 salary to your monthly fee income and concluded it fits. It does fit, in the way a sofa fits through a door if you don't count the doorframe.

What leaves your account every month, or lands as a bill you can't defer:

Get the current employer NI rate and secondary threshold from HMRC directly rather than from any blog, this one included — it moves, and an out-of-date percentage in your spreadsheet is worse than no percentage. Same for the pension: 3% is the minimum employer contribution, not a target.

As a rule of thumb we use when advising practices on this: the fully loaded monthly cost of a first junior hire lands somewhere around a third above their gross salary once the statutory stuff, the seats and the kit are in. Your number will be different. The point is that if the salary only just fits against your fee income, it doesn't fit.

Two of those lines are specific to us and get missed by every generic small-business guide.

Professional indemnity. Your PI premium is rated on fee income, work type and how many people are producing the work. Adding a member of staff is a material change and your insurer wants to know. Ring your broker before you advertise the role, not after you've made an offer — ask them, in plain terms, what the premium does if you go from one person to two and your fee income rises by the amount you're projecting. What we've found is that practitioners budget for the salary to the pound and then get a mid-term adjustment they hadn't modelled at all. It's not usually catastrophic. It is usually a surprise, and surprises are what kill small practices, not costs.

Software seats. Your subscription stack is per-seat and it scales in a straight line. A second person on Revit or ArchiCAD, plus whatever you render in, plus cloud storage and a second licence of anything you use for schedules or sheet issue, is a real annual number before they've drawn a line. Check the current list prices yourself rather than assuming the second seat comes cheap.

The Cost You Will Not Put in the Spreadsheet

Here's the one that actually bites.

A Part 1 assistant cannot be left alone with a building regs package. They shouldn't be left alone with a client email either. Everything they produce passes through you, and in the first few months a lot of it comes back for a second pass. That checking, redlining, explaining-why-the-wall-build-up-matters time comes directly out of your own fee-earning hours — and your hours are the most expensive thing in the practice.

In practice, a realistic first-quarter supervision load is something like an hour a day: setting the task properly in the morning, reviewing it properly before it goes anywhere. Call it five hours a week off your own billable total. At your own charge-out rate, that is very possibly the single largest line in the whole stack, and it's the only one that appears nowhere on a payroll report.

It does decay. By month six a decent assistant is producing survey drawings, existing/proposed sets and consultant coordination markups with a much lighter touch. But you have to plan for the front-loaded version, because the front-loaded version is the one that happens while your cash is tightest.

Now Do the Sum Backwards

Forget affordability for a second and calculate what the hire has to generate.

Take your fully loaded monthly cost. Divide it by the hours they can realistically bill. A full-time week is around 37.5 hours, but nobody bills 37.5 hours — deduct holiday, bank holidays, admin, CPD, office housekeeping, and the hours where they're learning rather than producing. In our experience a first-year Part 1 in a small practice is doing well to be genuinely chargeable 60–65% of the time, and lower than that in the first two months.

So: say the loaded cost is £3,000 a month and they're chargeable 60% of a 37.5-hour week. That's roughly 97 chargeable hours a month, which means every one of those hours has to recover a bit over £30 just to break even — before the practice makes a penny, before your supervision hours are accounted for, and before any of it turns into profit you can draw.

Then ask the only question that matters: do I have fee work at a rate that supports that, already signed, for the next nine months? Not enquiries. Not "the client's very keen." Signed appointments with a schedule of services and a fee stage programme.

If you're not confident your appointments hold up to that question, that's worth fixing before the hiring question — our piece on how to start your own architecture practice covers pricing and terms from the ground up.

Fee Stages Don't Pay Salaries. Months Do.

This is the structural problem with hiring in architecture and it's the reason generic "true cost of an employee" articles are useless to us.

Your income doesn't arrive monthly. It arrives when RIBA Stage 3 completes and the invoice clears, which might be week seven or might be week nineteen depending on whether the client's spouse decides the kitchen should face the other way. Salary arrives on the 28th regardless. You are converting lumpy, milestone-triggered, client-dependent income into a flat, non-negotiable, legally enforceable outgoing.

Which means the affordability test isn't annual. It's month by month. Plot your contracted fee stages on a calendar against the loaded monthly cost and look for the gaps. There is almost always a gap: planning determination periods, a technical design stage that can't start until the structural engineer reports, the client who goes quiet for six weeks in August. Those are the months that empty the account.

The buffer rule we teach: three months of fully loaded cost sitting in cash, untouched, and not counted as anything else. Not your tax reserve. Not your VAT. Not money earmarked for the new laptop. Three months of the employee's total cost, liquid, before you make the offer. If a project stalls and you've got a buffer, you have a problem. If a project stalls and you don't, you have an employment law problem, which is considerably worse.

And don't hire on the back of one big commission, however good it looks. One project is one client's change of mind away from disappearing. Hire against a pipeline of three or four appointments at different stages, so that a stall in one doesn't take the whole cost base with it.

Everything You Can Do Instead

A permanent full-time employee is the most expensive, least reversible option available. It's also, weirdly, the first one everybody considers. Here's what beats it, and when.

Subcontract to another architect or a freelancer. They invoice you, they carry their own PI, they use their own kit and licences, and when the work stops so does the cost. This is the right answer when your bottleneck is a specific deliverable — a technical package, a batch of visuals, a measured survey — rather than general capacity. The catch: you're paying a day rate that will look expensive against a salary, you get less control over programme, and if you're using the same freelancer every day of every week, take advice on employment status, because HMRC has views.

Part-time, two or three days. Halves the fixed liability, halves the supervision load, and lets you find out whether you actually enjoy managing someone before you commit. A lot of good Part 1s and Part 2s would take three days. This is the option we'd point most sole practitioners towards first, and it is chronically under-used because it feels like a half-measure. It isn't. It's a pilot.

Fixed-term contract tied to a project. If the pipeline is genuinely one big job across twelve months, a fixed-term appointment matches the liability to the income. Be straight with the candidate about what it is — people take fixed-term roles happily when they're named as such, and badly when they discover it later.

Share a hire with another small practice. Two sole practitioners, one assistant, split days and split cost. Administratively fiddly, and you need to agree who employs them and how PI and confidentiality work across two client lists. When it works it gives a junior far better exposure than either practice could alone.

A Part 1 year-out placement. Structured, time-bound, and both parties know it ends. It's also a contribution to the profession rather than just a resource decision, which is worth something.

The Test, Written Down

You're ready when, and only when, all of these are true at the same time:

• Signed appointments, with defined fee stages, covering the loaded cost for at least the next nine months
• Three months of loaded cost in cash, ring-fenced and not doing another job
• A month-by-month forecast that doesn't go negative in any single month
• Your broker has told you what PI does, and you've put that number in
• A charge-out rate and a realistic utilisation assumption that make the chargeable-hours sum work
• Actual named work for them in week one that isn't "sort the file structure"

Miss one and you're not being cautious by waiting. You're being solvent.

And if the sums say no but the workload says something has to give, the answer usually isn't a hire at all. It's raising your fees, or turning down the jobs you don't want at prices you'd resent, or subcontracting the bits you're slowest at. Turning work away is not a failure state. Hiring to avoid it, and then spending eighteen months working harder than before to cover a salary, very much is.