The real cost of your next hire isn't the salary

An agency wins a retainer that needs one more developer. The owner works out that the developer costs a certain amount per month, checks it against the retainer value, sees a margin, and posts the job. Three months later the margin has not appeared and nobody can quite explain where it went.

It went into the parts of hiring that never make it onto the spreadsheet.

The unbillable weeks

Start with time. Writing the post and screening applicants is real work done by someone senior. First-round calls, second-round calls, a take-home task somebody has to review properly. Then an offer, a negotiation, and a notice period you do not control. Six to ten weeks from decision to first day is normal, and none of it is billable.

Then onboarding. Even a strong hire is not productive on day one. They need access, context on the client, an understanding of how your team actually works versus how the wiki says it works. The people giving them that context are your most senior people, which means the cost of onboarding is paid in exactly the hours you can least afford to lose.

Call it four weeks to meaningful output, which is optimistic. You are now roughly eleven weeks from the decision, and you have been paying a salary for the last four of them.

The costs that are not time

On top of the salary itself, the recurring additions are easy to underestimate:

• Employer taxes and statutory contributions
• Equipment, usually front-loaded in month one
• Software seats — design tools, repos, project management, each billed per head
• Paid leave, public holidays and sick days, which reduce billable capacity without reducing cost
• Recruiter fees, where used, typically a meaningful slice of first-year salary

None of these are controversial. They are just rarely in the number the owner compared against the retainer.

The risk nobody prices

The real exposure is not the cost. It is the shape of the commitment against the shape of the work.

A salary is a flat monthly line that does not care what your pipeline is doing. Agency revenue is lumpy. You win three projects in April and none in July. The hire you made for April is still on payroll in July, and in July they are an expensive reminder that you forecast optimistically.

You are matching a fixed cost against a variable revenue line and hoping the averages work out. Sometimes they do.

And if the hire does not work out — wrong fit, wrong level, or simply someone who leaves after seven months — you do not just lose the time. You restart the eleven weeks, and you do it while already behind on the work that prompted the hire.

When hiring is still the right call

We are a delivery bench, so take this with the appropriate pinch of salt — but hiring is often correct, and we will say so on a call.

Hire when the need is permanent and central. If the skill is what your agency is known for, it belongs in-house. If the role is senior enough to shape how work gets done rather than just execute it, it belongs in-house. If you have twelve months of visible, committed demand for that capability, the arithmetic above stops being frightening.

Where it stops making sense is the eight-week project in a discipline you touch twice a year. Nobody should carry a permanent salary to service occasional demand. That is the gap a bench exists to fill, and it is a narrower gap than most of our sales material implies.

A simpler way to decide

Before posting the role, answer two questions honestly. How many months of committed work can you name for this person right now — not hoped-for, committed? And if that work finished tomorrow, how long could you carry the salary while finding more?

If the first number is large and the second is comfortable, hire. If the first is small or the second makes you wince, you have a project, not a role. Those need different solutions.