Finance Said Our Dubai Dev Hires Cost AED 22,000 Each: the 7 Steps That Found AED 38,220

The conversation that started this was short. A budget holder asked what it costs us to hire a developer, I said about AED 22,000 because that was the number in the plan, and a colleague who had been in the interview loop all month raised an eyebrow. She was right to. The AED 22,000 was the sum of the invoices we receive, which is a measure of what we pay suppliers rather than what hiring costs us. Rebuilding it properly across the last fourteen hires took about five days of careful work and produced a representative figure of AED 38,220, with a blended average near AED 38,600. Nothing new was being spent. We simply had not been counting most of it. Here is the method, in the order that keeps it defensible.

Step 1: Fix the definition before you collect a single number

Every failed attempt at this metric fails here, by starting a spreadsheet before agreeing what belongs in it. Adopt the standard structure: external costs plus internal costs, divided by the number of hires in the period. That is the shape of the published cost-per-hire standard developed by SHRM and ASTM, designated ASTM E3032, and the reason to use it rather than your own invention is that it lets you compare your figure with someone else’s and with your own figure next year.

Then settle the four line items people argue about, in writing, before you look at any data: immigration and onboarding administration, referral bonuses, the cost of candidates you rejected, and recruiter salary. Our choices were to include immigration and onboarding, include referral bonuses as an external cost, include rejected-candidate interviewer time, and exclude recruiter base salary on the grounds that it is fixed overhead rather than a per-hire cost. You may reasonably choose differently on all four. What you may not do is decide after seeing the numbers, because at that point the inclusion list becomes a way of reaching a preferred answer.

Fix the denominator with the same care. Count hires who actually started, not offers accepted, and use the same window for costs and hires. We used eleven months and fourteen hires. Write the window next to the figure permanently, because a cost per hire with no stated period invites exactly the misreading that started this exercise.

Step 2: Pull external costs from the ledger, not from memory

External cost is the easy half and still worth doing properly. Ask accounts payable for everything coded to recruitment in the window rather than reconstructing it from recollection, because the small recurring items are the ones memory drops. For our representative hire, an agency placement of a mid-level backend engineer, external cost came to AED 20,900: an AED 18,000 agency fee, AED 1,500 of allocated job board and sourcing platform charges, AED 500 of allocated assessment tooling, and AED 900 of verification and reference checking.

The allocation of annual subscriptions is where this step goes wrong. A sourcing platform costing AED 21,000 a year is not AED 21,000 against one hire, nor is it AED 1,500 against every hire by default. Divide it across the hires that channel actually produced in the window. Ours produced five, so it carried AED 4,200 each in the direct-sourcing channel and nothing in the agency channel. Getting this wrong is what makes in-house sourcing look free, which in turn makes step 7 produce a conclusion you will regret acting on.

Step 3: Price interviewer time at a fully loaded rate

This is the step that moved our number most, and it requires two decisions. First, a loaded hourly cost for the people in the loop. Take annual salary, add roughly twenty-five percent for employer costs, and divide by productive hours rather than calendar hours. For our engineering panel that produced a blended AED 240 per hour. Use a blended rate across the panel rather than per-person rates: the precision gain is negligible and the political cost of publishing what each engineer costs per hour is not.

Second, count every hour honestly, including the ones that feel like they do not count. Per hire we counted 26 hours of engineering time, which at AED 240 is AED 6,240. That total includes preparation, the interviews themselves, scorecard writing afterwards, and the debrief. It also includes time spent on the candidates we did not hire, which is the single largest omission in most attempts at this metric. If three finalists reach a technical loop and you hire one, two thirds of that interviewing cost belongs in the figure. Leaving it out does not make it cheaper, it makes your funnel look efficient when it is in fact expensive, and it removes the incentive to screen better earlier. A structured loop is the main lever here: the work in building a technical interview scorecard and in running a pair programming interview pays for itself in hours you stop spending.

Step 4: Count the coordination and admin hours nobody logs

Beyond the interview loop sits a layer of work that is real, substantial and completely invisible in any system of record. Scheduling across calendars and time zones, chasing feedback, assembling the offer, the approval chain, and then the Emirates onboarding sequence. We counted 14 hours per hire of recruiter, hiring manager and operations time at a loaded AED 220 per hour, which is AED 3,080.

Then the UAE-specific block: the employment visa, the medical, the Emirates ID, and the administration that surrounds all three. We account for that at AED 6,000 per hire including the internal time to run it. Your figure will differ with free zone, role and whether you use a service provider, and it is worth getting from your own records rather than from an article. What matters is that it appears somewhere, because it is a cost that does not exist for a hire in many other markets and it is one of the reasons a Dubai figure should not be benchmarked naively against a European one. Much of this time is onboarding rather than recruiting, which is why step 5 exists, and why a good onboarding programme shows up in this metric as well as in retention.

Step 5: Decide where visa and onboarding sit, and write the decision down

This step is a sentence of documentation that saves an argument every quarter. Immigration and onboarding administration are defensible either way. Excluded, they are employment costs incurred after the candidate said yes, and your figure stays comparable with published benchmarks that usually exclude them. Included, they reflect that no hire exists in the Emirates until that sequence completes.

We include them, and we say so on the same slide as the number, in one line: cost per hire includes employment visa, medical, Emirates ID and associated internal administration; excludes recruiter base salary; eleven-month window; fourteen hires. That line is the difference between a metric and a talking point. It also makes a future comparison honest, because the first question about any year-on-year movement should be whether the definition moved too.

Step 6: Keep ramp and vacancy as separate metrics

There is a strong temptation at this point to keep going, add the productivity ramp and the cost of the empty seat, and produce one big number. Resist it, because it breaks the metric. Cost per hire exists to tell you whether your acquisition process is efficient. Ramp is a function of onboarding quality and codebase complexity. If you fold ramp in, the combined figure moves when your onboarding changes, and you can no longer read it as a signal about sourcing.

Report both, separately, under different names. We call the second one cost of the seat. For the same representative hire: ramp of roughly 90 days at an average effectiveness we estimate at 45 percent is about 1.65 months of salary foregone, which at AED 22,000 a month is roughly AED 36,300. Added to AED 38,220 that gives a cost of the seat near AED 74,520. That is the number for a headcount business case, not for a sourcing review.

On vacancy cost, we deliberately publish no figure, and I would encourage the same restraint. Valuing an unfilled engineering seat requires assuming what the delayed work would have been worth, which for most teams is an estimate dressed as a measurement. We report days to fill as a plain number and let the budget holder apply their own judgement about what the delay cost. If you do want a delivery-side view of the same question, DORA metrics are a more honest instrument than a monetised vacancy estimate.

Step 7: Rebuild the figure per channel and let it change sourcing

The blended average is the least useful output of this whole exercise. AED 38,600 across fourteen hires tells you almost nothing actionable, because it averages channels whose economics differ by a factor of two and a half. Run the same calculation separately for each channel, with the subscription allocation from step 2 applied correctly, and the picture becomes a decision.

Two conclusions came out of this for us, and only one of them was obvious. The obvious one: six agency placements at AED 52,000 is where the money is, and some of those searches did not need an agency. The less obvious one: referral at AED 21,100 is not a mandate to push referrals harder. Referral volume is capped by the reach of the team you already have, and a team that hires mostly through its own network tends to reproduce its own profile, which is a real cost that never appears in a cost-per-hire table. We moved two planned searches from agency to direct sourcing, left the referral programme as it was, and kept agencies for the two genuinely hard profiles.

Pair the channel figures with retention at twelve months before concluding anything. A channel that is cheap per hire and loses people inside a year is the most expensive channel you have, and cost per hire on its own will never show you that. The upstream fixes that improve both numbers are usually the same ones: a job description that filters accurately and reference checks that are actually conducted cut wasted interview hours and bad starts at the same time. If you want to model the ongoing cost of a team rather than the acquisition cost of one person, our team cost calculator covers the other half of the question.

Frequently asked questions

What is the standard cost-per-hire formula?

External costs plus internal costs, divided by the number of hires in the same period. That is the structure of the cost-per-hire standard developed by SHRM and ASTM, designated ASTM E3032, and using it rather than your own invention is what makes your figure comparable. External costs are what leaves the company: agency fees, sourcing platform charges, assessment tooling, verification checks. Internal costs are what you spend without an invoice: interviewer hours, recruiter and hiring manager hours, coordination and onboarding administration. Count hires that actually started, in the same window as the costs, and state the window next to the figure.

Should visa, medical and Emirates ID costs be inside cost per hire?

It is a judgement call and the only wrong answer is leaving it undocumented. Excluding them keeps your figure comparable with published benchmarks, which usually treat them as employment rather than recruiting costs. Including them, which is what we do in the Emirates, reflects that no hire exists until that sequence completes and that at roughly AED 6,000 per head including administration time it is material. Pick one, write it on the same page as the number, and keep it stable between periods.

Why keep the productivity ramp out of cost per hire?

Because it answers a different question. Cost per hire tells you whether acquisition is efficient. Ramp is driven by onboarding quality and codebase complexity, so folding it in produces a figure that moves when onboarding changes and can no longer be read as a signal about sourcing. Report both under different names. We use cost per hire for acquisition and cost of the seat for acquisition plus ramp, and the second is what belongs in a headcount business case.

Which channel gives the lowest cost per hire in Dubai?

In our fourteen-hire sample, referral was lowest at roughly AED 21,100, direct in-house sourcing sat at roughly AED 33,000, and agency placements reached roughly AED 52,000, against a blended average near AED 38,600. That ranking is not an instruction to move everything to referrals: referral volume is capped by your team’s reach and tends to narrow the profile you hire, direct sourcing consumes recruiter capacity, and agencies earn their fee on genuinely hard searches. Treat it as a budget reallocation question, and always read it alongside twelve-month retention.

The formula structure follows the SHRM and ASTM cost-per-hire standard (ASTM E3032). All figures in this article are from our own records across fourteen engineering hires in Dubai over an eleven-month window, calculated with the inclusion list stated in step 5. Your loaded hourly rates, subscription allocations and onboarding costs will differ, and should be taken from your own ledger rather than from this example.