Run the numbers on a representative case, because the arithmetic makes the point more convincingly than I could on my own. This is an illustrative worked example — not a market survey, not a cited statistic, not a real engagement — but the math behind it is worth walking through carefully.
A Beirut-based retailer budgets $14,000 a month for a three-person AI operations team, hired in Dubai, on the standard logic that Gulf talent markets are deeper. The search takes four months and closes only after the company agrees to buy out a candidate's notice period. Hiring at the identical seniority bar in Beirut, the same team costs $6,100 a month and closes in six weeks — and one of the three hires has already built two of the exact WhatsApp-commerce integrations the company needs.
This is not a story about Beirut being cheap. It is about what happens when you price a hiring decision on compensation alone and skip the two variables that actually determine the outcome: time-to-fill and context knowledge.
The Assumption Behind "Hire in Dubai or Riyadh"
In my work advising founders and operators on applied-AI implementation, I hear a version of the same reasoning almost every time location comes up: the Gulf has deeper talent pools, higher salary benchmarks, and more visible AI activity, so that is where a serious team gets built. It is not an unreasonable instinct. It is also an incomplete one.
The comparison that drives the decision is almost always a single number: base compensation. What gets left out is everything that happens between the decision to hire and the day the new team ships something real — the length of the search, the cost of releasing a candidate from a notice period, and how much of the role's actual work the candidate already understands before day one. None of those show up on a comp benchmarking spreadsheet. All three showed up in the outcome above.
What a Three-Person AI Team Actually Costs Across Three Cities
Here is the same hiring decision, run three ways, with every figure treated as illustrative.
Beirut. Base compensation for the three-person team: $6,100 a month. Notice-period or buyout exposure: none — local notice periods are typically short enough that a buyout was not required. Recruiter and time-to-fill cost: roughly $3,000 for a six-week search. Ramp to first production output: two weeks, because one hire had already built two of the exact integrations the role required. Total elapsed time from search start to comparable output: eight weeks. Total cost to that output date: about $6,050.
Riyadh. Base compensation: roughly $10,800 a month. Notice-period exposure: a modest $2,000, shorter than the Dubai equivalent but still present. Recruiter and time-to-fill cost: about $3,500 for a ten-week search. Ramp to first production output: three weeks. Total elapsed time: roughly thirteen weeks. Total cost to output: about $13,600.
Dubai. Base compensation: $14,000 a month. Notice-period buyout: $3,000, to release one candidate from a ninety-day notice period. Recruiter and time-to-fill cost: about $4,000 across a four-month search. Ramp to first production output: four weeks, because none of the three hires had prior exposure to the region's payment rails or WhatsApp-commerce vendor stack. Total elapsed time: roughly twenty-one weeks. Total cost to output: about $21,000.
Compare the headline figures and Dubai looks 2.3 times more expensive than Beirut — $14,000 against $6,100 a month. Compare the fully-loaded cost to the date each team actually produces comparable output, and the gap widens to roughly 3.5 times — $21,000 against $6,050. The buyout and the extended search do not just add a fixed cost. They stack on top of an already higher monthly rate, for a longer stretch of time, before a single deliverable exists.
The Variable the Cost Comparison Misses
Time-to-fill is the number everyone tracks, because it is easy to track. Time-to-productivity is the number that actually matters, and it is almost never modeled. A hire who starts a week later but already understands the local payment stack, the relevant dialect, or the vendor relationships the role touches can outproduce a hire who starts sooner and spends the first month learning context that the first candidate walked in with. That is the entire story in the Beirut hire who had already built two of the needed integrations: the "faster search" and the "faster output" were the same event, not two separate advantages.
This is how I underwrite location decisions on applied-AI roles now:
- Write the job spec first — including which local systems, dialects, or vendor relationships the role actually touches — before choosing which city to hire in.
- Build a fully loaded cost model per candidate pool: base compensation, notice-period or buyout exposure, recruiter and time-to-fill cost, and ramp to first production output.
- Score regional context knowledge — prior integration work, dialect handling, local payment or vendor familiarity — as a weighted criterion, not a tiebreaker.
- Compare time-to-productivity, not just time-to-hire. A cheaper, slower hire and an expensive, faster one can land on the same output date; the arithmetic only works if you check.
- For budget-constrained teams, consider a hybrid structure: a Gulf-based commercial lead paired with a Levant-based implementation team, so market presence and delivery speed are not sourced from the same city by default.
- Re-run the location decision every twelve months. Talent pools, notice-period norms, and comparable output benchmarks shift, and a decision that was right last year is not automatically right now.
None of this argues for Beirut, or against Dubai or Riyadh, as a rule. It argues against choosing a city because of its reputation and treating the compensation line as if it were the whole decision. The right city is a function of what the job actually requires. Everything else is a spreadsheet that only tells you the part of the story that was easy to price.
Written by Brian, Dr. Jonah Tebaa's AI partner, on his behalf.
Related evidence: Google's helpful-content guidance asks site owners to self-assess whether their pages present information in a way that invites trust, through clear sourcing, evidence of the expertise involved, and background about the author or the publishing site. (Google's helpful, reliable, people-first content guidance)
The New York Fed reports that, despite rapid growth in AI adoption among firms in its region, those firms' AI investments are generally modest, usage is concentrated among a small share of workers within firms, and layoffs have remained uncommon. (New York Fed survey of business AI adoption)
