The market that brought you here is two hundred markets, and it has already shrunk once

Nearly every argument for moving a business to this country arrives folded inside one sentence about the people in it. Large. Diverse. Growing. International. More than two hundred nationalities. A ready-made market of millions who already speak your language and already buy what you sell.

The sentence is not invented. It has a source, and the source is a government page anybody can open. What almost nobody does is open it — because the sentence reads like a fact about a place, and facts about places do not seem to need checking.

Read it, and the page turns out to say four separate things. One of them is the sentence. The other three are not, and each of the three lands somewhere specific in a business plan written before the founder arrived.

Line one: the headline total is five years old, and the split underneath it is fifteen

The UAE Government's own fact sheet, citing the Federal Competitiveness and Statistics Centre, gives the total population — nationals and expatriate residents together — as 9,282,410 in 2020.

The Ministry of Foreign Affairs, on its own facts and figures page, gives 9,304,277, for 2017.

Neither is wrong. They are two official pages, published by two federal bodies, quoting the same statistical authority at two different moments, and neither has been carried forward. That is the first thing worth knowing about population data here: the authoritative series is real, it is published, and the versions of it reachable from the front door of the government are not current.

The nationality split is older still. The most recent breakdown the Ministry publishes is dated 2010: 947,997 nationals and 7,316,073 non-nationals. Add them and the total of that year comes to 8,264,070, which puts the national share at just under twelve per cent — our arithmetic on their two figures, not a published percentage.

So the shape of what is available is this. A total from 2020. A different total from 2017. A composition from 2010. The figure most business plans lean on hardest — how many people are here, and who are they — is assembled from three vintages that do not meet.

None of that argues the market is small. It argues that anyone quoting you a current population figure with a decimal point on it got it somewhere other than these pages, and you should ask where. The route to numbers that are actually current — the federal open data portal, the emirate's own statistics establishment, and the fact that some of it arrives on request rather than as a download — is the subject of Dubai publishes your competitors and nobody publishes your customers.

Line two: the number went down

Set the two federal totals side by side in the order they were measured, and the direction is not the one the sentence promises.

9,503,738 in 2019. 9,282,410 in 2020.

A fall of 221,328 people. About two point three per cent of the country, in a single year, in the government's own published series.

We are not going to tell you why, because the honest answer is that the fact sheet does not say and we did not find a methodology note covering the break. The obvious candidate is the year it happened in. It may be partly that, partly a change in how the count was taken, partly neither.

The cause is not actually the important part. The important part is that the mechanism which permits a fall of that size is not a one-off event, and it has not gone anywhere.

Almost everyone in this country is here on a permission that is attached to something else — an employer, a company, a licence, a spouse, a parent. Presence is not a right anybody holds independently; it is a derivative of a file that somebody else maintains and that has a renewal date on it. When the thing underneath goes, the permission goes, and after a grace period the person goes too.

That is the same structure described from the founder's side in a trade licence is not a Golden Visa and from the household's side in moving the household is the longer project. What the population series adds is the aggregate view of it: a country where residence is conditional is a market that can contract quickly, because nothing has to be built or demolished for it to happen. A hundred thousand people can leave without a single shop closing.

No European or North American market a founder is likely to be arriving from has that property. Plans imported from those markets do not have a line for it.

Line three: in the official count there are two point three men for every woman

The same 2020 fact sheet splits its total: 6,468,460 males and 2,813,950 females.

Women are a shade over thirty per cent of the recorded population. This is not a subtle skew or a rounding artefact. It is the single most consequential line on the page for anybody selling to consumers, and it is the line that never appears in a benefits summary.

What it does to a plan depends entirely on what the plan sells.

If your product is bought by households, the household composition behind that ratio is not the one your home market taught you. A large share of the male majority is here without the family the product was designed around.

If your product is bought predominantly by women, your addressable market is roughly a third of the headline you were quoted, before any other filter — income, age, location, language — has been applied. A market-sizing exercise that starts from the total and applies a category conversion rate has already overstated the answer by a factor of three.

If your product is bought predominantly by men, the reverse holds and nobody will tell you either, because the sentence you were sold was silent on sex in both directions.

The correction is cheap. It is one multiplication, done before the pricing model rather than after the first quarter. What makes it expensive is that the ratio is not intuitive, so it never gets questioned — and a wrong denominator does not announce itself. It just produces a revenue forecast that misses for reasons the founder attributes to marketing.

Line four: more than two hundred nationalities is a warning, not a market description

The Ministry's page states that more than 200 nationalities live and work in the UAE, and the government portal names the largest communities in order — Indians first, then Pakistanis, Bangladeshis, other Asians, Europeans and Africans. It publishes that ordering without a population figure attached to any single community, which is why none appears here.

This line is the one most often read as pure upside, and it is genuinely an advantage — for a business whose customers are other businesses, or whose product is indifferent to language, culture and habit.

For a consumer business it is the opposite of what it sounds like. Two hundred nationalities is not one market of nine million. It is a large number of segments that shop in different languages, at different price points, through different channels, on different pay cycles, and that mostly do not read the same advertising. A campaign that would reach a national market at home reaches a slice here — and the slice is not defined by the demographic categories the campaign was built on.

Two practical consequences follow, and both are cheap to act on at the planning stage and expensive later.

The first is that your customer acquisition cost is a per-segment number, not a national one. Winning the first segment tells you very little about the second, because the channel that worked does not carry across.

The second is that "diverse" and "addressable" are different words. The population is genuinely international. Your product is not automatically international because it is standing next to it.

What the sentence claims, and what the page it came from actually says

Nothing in the right-hand column contradicts the left-hand column outright. That is precisely the difficulty. Every line on the left is a defensible reading of a line on the right, with the awkward half removed — and the awkward halves are where the planning decisions live.

The same fact reaches your customers and your staff at the same time

There is one more consequence, and it is the one founders are least prepared for because it arrives from two directions at once.

The population whose presence is conditional is also the population you hire from. The permit that makes someone a customer is frequently the same instrument that makes someone an employee — and when your own company sponsors it, the conditionality moves onto your file. A resignation here is not only a resignation. It is a cancellation, a grace period, a status change or a departure, and a set of end-of-service obligations that grew quietly with tenure. The mechanics, and what the first hire actually opens, are in hiring the first person opens a file the licence did not.

So the churn in the market and the churn in the payroll are not two risks. They are one property of the country, showing up on both sides of the profit and loss account. A plan that models staff turnover at home-market rates and customer retention at home-market rates has made the same mistake twice and will find it twice.

Against that, the honest counterweight: the same conditionality is why hiring is fast, why notice periods are short, and why a business can scale a team here in weeks rather than quarters. The flexibility and the fragility are the same fact read from opposite ends. What you cannot do is budget for one and not the other.

Where this is the wrong thing to worry about

This page is a corrective, and a corrective applied to the wrong business is just an obstacle.

Your customers are businesses, not people. Almost none of the above bites. Company counts, sector concentration and licence data answer your question, and they are public and current in a way the population series is not. Start at the federal lookup, not at the fact sheet. Reading a sector before you enter it — the gates on it, and which one binds first — is the subject of choosing a business sector in Dubai.

Your customers are not in the UAE at all. A very large share of companies formed here sell abroad and use the country as a base, a banking relationship and a time zone. For them the domestic population is scenery. The relevant questions are the ones in what the UAE actually gives you and in your UAE company is a dollar business, and none of them turns on who lives here.

You are already trading here and have your own data. Twelve months of your own transactions beat every figure on this page, and beat them by a wide margin. National statistics are for the period before you have customers. Once you do, they are a curiosity.

You have been shown a current, precise figure by somebody who cited its publisher. Then use theirs. This page is not an argument that the numbers are unknowable — the emirate-level series exist and Dubai maintains a live population clock alongside a separate count of the people present during the working day. It is an argument against accepting a figure with no publisher attached to it.

You are treating the male majority or the fall as a reason not to come. Neither is. A market of this size, this open and this well connected does not become a bad market because its composition differs from the one you left. It becomes a market that has to be sized properly. The failure this page is written against is not optimism; it is arithmetic performed on the wrong denominator.

What we ask a founder before any of this matters

Four questions, and they take longer to answer honestly than to read.

1. Who buys this — a person or a company? The whole page applies to one answer and almost none of it to the other.
2. If a person, what proportion of your buyers at home were women? Then apply the published ratio and look at the result before you look away from it.
3. In how many languages does your product have to work before it has a market here? If the answer is more than one, that is a cost line, and it belongs in the first budget rather than the second.
4. What happens to your revenue if five per cent of your customers leave the country in a year? Not a forecast, and deliberately set above the two point three per cent the published series actually records. A stress test is only useful above the observed case, and the reason to run this one at all is that the country has an observed case.

A brief that answers those four is worth more than any market study bought afterwards, and it is the half of the work a consultancy genuinely cannot do for you — the argument in the brief is the half a consultancy cannot do for you.

Frequently asked questions