Cost-per-lead tables for the UAE market circulate until they acquire the authority of a standard, usually without anyone being able to say who measured them or what they counted. Here is what a benchmark can honestly tell you, and how to build the only baseline that matters — your own.
A benchmark only means something if you know what it counted, where, and when. Most of the ones circulating here answer none of the three.
Every few months a table of UAE marketing costs turns up in a deck or a LinkedIn carousel: impressions priced by industry, cost per lead by vertical, a target return on ad spend for e-commerce. It gets screenshotted, pasted into somebody else's pitch, quoted in a board meeting, and within a quarter it has acquired the authority of a standard. Ask where it came from and the trail usually ends at another deck.
Meanwhile the gap between two advertisers in the same category in Dubai is routinely wider than the gap between two categories. Offer, creative, audience size, landing page, season, the state of the auction in the week you happened to buy — and above all what each advertiser decided to call a lead — move the number far more than the industry label does. Two accounts can report costs that look nothing alike and both be run well.
None of this makes benchmarks worthless. It makes them a sanity check rather than a target: useful for noticing that you are out by a margin nobody can explain, useless for deciding what your cost per acquisition ought to be. The number that decides whether a campaign is working comes from your own margin, and it is arithmetic rather than research.
If a table cannot survive these six, it is not evidence. It is a slide with numbers on it.
A named organisation publishing its own method, or an anonymous aggregate nobody can inspect. If the sample is described as somebody's private ad accounts, treat the table as an anecdote with formatting.
A lead is not a unit of measurement. A cost per lead built on click-to-WhatsApp taps and one built on qualified, contactable enquiries are two different measurements, and neither predicts the other.
Auction prices here move with the calendar — Ramadan and Eid, the summer trough, retail festival periods, and the property and travel cycles that sit on top of them. A quarter-long average hides all of it.
The GCC is often reported as one block. Dubai, Abu Dhabi, Riyadh and Kuwait do not price alike, and a figure converted from dollars at an unstated rate is one more unverifiable step away from your own budget.
The same campaign reports different conversion counts under a seven-day click window, a one-day view window, or last-click in analytics. Change the model and the cost per conversion moves without the account changing at all.
A single average tells you nothing about the spread it came from. Ask for the range and the number of accounts behind it. If neither is available, the average is not something you can plan against.
The same six questions work on a report you commission, on a competitor case study, and on anything I tell you. If I quote a figure and cannot show you where it came from, apply the first question to me.
Where a platform defines a threshold itself, the figure is checkable. That is the standard for anything quoted on this site.
An earlier version of this page carried a table of UAE cost benchmarks — media costs, cost per lead and target return on ad spend by industry — sourced to my own aggregated ad account data. I have taken it down. The sample was never defined, the accounts behind it are not mine to publish, and a figure you cannot audit is decoration rather than evidence. What follows is the opposite case: a set of thresholds Google publishes and documents, which anyone can check against their own site today.
| Core Web Vital | Good | Needs improvement | Poor |
|---|---|---|---|
| Largest Contentful Paint (LCP) | 2.5 s or less | 2.5 s – 4.0 s | Over 4.0 s |
| Interaction to Next Paint (INP) | 200 ms or less | 200 ms – 500 ms | Over 500 ms |
| Cumulative Layout Shift (CLS) | 0.1 or less | 0.1 – 0.25 | Over 0.25 |
Source: Google's published Core Web Vitals thresholds, documented on web.dev. A page is assessed at the 75th percentile of real visits, with mobile and desktop measured separately — so the score reflects your slower visitors rather than an average that flatters you.
These are worth knowing in this market because most of the traffic I buy here arrives on a phone, often on mobile data, and a landing page that fails an interaction threshold loses conversions you have already paid for. It is a benchmark you can act on the same week, which is more than can be said for a cost table. The technical side of that work sits under SEO.
Five steps. None of them require anyone else's data, which is precisely why they work.
Write down the exact event you are costing, where it is recorded, and who agrees that it counts. Not a lead — a submitted form with a reachable phone number, or a WhatsApp conversation that got a reply, or a booked appointment that showed up. Almost every argument about whether a cost per lead is good turns out, on inspection, to be an argument about what was being counted.
Pixels and server-side events firing on the right action, a UTM convention nobody is allowed to improvise on, and one place where the number is considered true. Where the platform and analytics disagree, decide in advance which one the reporting uses and why. A baseline built on top of a tracking layer nobody trusts inherits that distrust.
Launch weeks are not representative: the auction is still learning, creative is untested, and audiences have not settled. Give it a stable run at a stable budget before you read anything off it. Judging cost per acquisition after a few days rewards whoever is loudest about early numbers rather than whoever is running the account well.
Then segment: by campaign, placement, creative angle, device and emirate. Your own trend line answers the question a borrowed table cannot — is this getting better or worse, and which part of the account moved it. A competitor benchmark cannot tell you that even when the number happens to be accurate.
The number that governs the budget is the most you can pay for a customer and still make money — derived from gross margin, close rate and repeat purchase, not from an industry table. Once that ceiling exists, a market average becomes a curiosity rather than a target, and you can tell in week three whether a channel is viable.
A first campaign has no baseline by definition. That is an argument for measuring carefully, not for borrowing somebody else's average.
The engagements described on work — property, travel, beauty e-commerce, automotive detailing — are set out as scope rather than as results for exactly this reason: a number from one of those accounts would tell you nothing reliable about yours, and publishing it would only add another untraceable figure to the pile.
Measurement is not a separate discipline from the work. It is the part that decides whether the rest of it can be judged.
The tracking layer this piece describes is the first thing I fix on a paid account, because optimisation is only ever as good as the data underneath it.
See how paid work runsCore Web Vitals are a search concern as much as a conversion one. Technical health is where an SEO engagement starts rather than where it ends.
See how search work runsCost per acquisition, ROAS and payback, e-commerce margin: the arithmetic behind your own ceiling, in the same working files I use on projects.
Open the tools