Salary benchmarking in the UAE: a working method for comp teams

Salary benchmarking in the UAE fails in predictable ways: a title matched to the wrong level, a basic salary compared against a total-cash figure, a band trusted because it appeared in a PDF. The method below is the one the Tenure dashboard automates, written out so you can run it deliberately, whether you use the tool or not. Five steps, in order.

Step 1: fix the unit before you compare anything

Decide what number you are benchmarking, and hold it constant on both sides. In the UAE the practical unit is total monthly cash: basic salary plus fixed allowances (housing, transport, and any other guaranteed monthly payment), excluding bonus and equity. It is the unit offers are negotiated in, and it sidesteps the basic-versus-allowance split, which employers set differently for gratuity reasons rather than market ones.

The most common benchmarking error is comparing your basic salaries against a market total-cash figure, which makes everyone look 20 to 40 percent underpaid. If your HRIS stores basic and allowances separately, sum them before you start. For where the line sits, see what counts as basic salary.

Step 2: match levels, not titles

Gulf titles are inflated and inconsistent. The same person is a Vice President at one bank, a Director at another, and a Senior Manager at a third. Matching on title strings will place a third of your team on the wrong rung before you have compared a single number.

Match each person to a rung on a canonical ladder instead: what they own, who they report to, and what the market calls that scope in their sector. The Tenure Pay Index normalises every band to a canonical role and level per sector for exactly this reason, and the sector pages show each ladder rung by rung, so you can place a role by reading the level definitions rather than guessing from the title.

Step 3: demand a source count for every band you use

A benchmark is only as good as the evidence behind it, and the evidence varies row by row within the same dataset. Before you rely on a band, ask how many independent data points support it. If the dataset cannot answer per row, you cannot weight your own conclusions.

The working thresholds Tenure publishes: below three sources a band is not shipped at all; at three or four it reads as directional; at five to nine it is suitable for setting pay ranges; at ten or more it will carry an offer decision or a leadership paper. The full policy is on the methodology page. Whatever dataset you use, write your own thresholds down; a threshold that lives in one analyst’s head is not a method.

Step 4: read percentiles, not averages

An average tells you almost nothing actionable, because UAE pay for the same role splits sharply by employer tier, and an average blends tiers into a figure that fits none of them. What you want is position: which percentile each person sits at within the market distribution.

Percentiles turn benchmarking into decisions. Someone at P30 in a role you cannot afford to lose is a retention risk with a number on it. Someone at P80 is either a deliberate premium or a quiet overpay. A whole function clustered below P25 is a structural problem, not a person-by-person one. The Pay Index publishes P10, P25, median, P75, and P90 on every band so that placement is a lookup, not a modelling exercise.

Step 5: re-run on a schedule, not on a resignation

Most teams benchmark when someone resigns, which is the most expensive possible trigger. The market has already moved, the counter-offer is already framed, and the rest of the team is watching. Benchmarking quarterly, against a dataset that itself refreshes quarterly, turns the same work into an early-warning system: you see the gap opening before the resignation email names it.

Date every run and keep the results. A benchmark without a date is a liability twelve months later, when someone quotes it in a comp review as if it were current.

Running the method

You can do all of this manually against the published sector pages. The faster paths:

One role. Run the free benchmark. Type the role title, level, country, and salary, and it returns the percentile that salary sits at against the live Pay Index, plus the medians for neighbouring rungs. No signup.

A whole team. The dashboard’s team benchmarking takes a CSV of your roster, matches each row to the closest sector, level, and market band, and returns the percentile per person, with the below-P25 retention risks ranked first. Uploads are isolated to your organisation and never feed the published dataset.

Both run on the same data: verified primary sources across 12 sectors in UAE and Saudi, source counts on every band, refreshed quarterly. Full access is $299 a month, flat, with a 14-day money-back guarantee on the first payment.