There's a lazy version of the Dubai e-commerce growth story that gets repeated every quarter: the market is maturing, growth is slowing, the easy money is gone. Talabat's third-quarter results, published yesterday, quietly demolish it. Group GMV hit USD 3.0 billion, up 16% year on year, and management raised full-year guidance for the second time this year.
That is not a mature market flattening out. That is a market still finding new customers, more frequent orders and bigger baskets, eleven months into 2026. If you run a business in Dubai that touches commerce, delivery, payments or the software behind any of them, this is your quarter to read carefully.
The three lines that actually matter
Everyone quotes GMV. The interesting part is where it came from.
GCC GMV reached USD 2.3 billion, up 9% year on year, and now represents 75% of the total, down from 80%. Meanwhile non-GCC GMV grew 43% to USD 736 million. Read together, those two lines say something specific: the home market is still growing at a healthy clip, but the company's next leg of growth is being built outside the Gulf.
Revenue grew 18% to USD 1.2 billion, which is faster than GMV. That's the second line worth your attention. Revenue outgrowing GMV means the mix is shifting toward higher-value activity, mainly own-grocery revenue through talabat mart, partly offset by lower commissions as grocery and retail take a bigger share of orders.
The third line is the honest one. Adjusted EBITDA came in at USD 150 million, 3% lower year on year and 5.0% of GMV, down from 6.0%. Net income fell 15% to USD 102 million, or 3.4% of GMV. Free cash flow, by contrast, rose 41% to USD 141 million with a 94% cash conversion ratio.
Growth was bought, deliberately, with margin. And the company still ended the quarter with USD 711 million net cash.
Growth is a product decision, not a discount
The cleanest detail in the release is almost a footnote: roughly USD 9 million of planned customer incentives in Food didn't need to be spent, because competition turned out less intense than budgeted.
That tells you the growth isn't purely promotional. Talabat is holding customers with selection, delivery reliability and value, and it's spending its USD 120 million Board-approved investment programme on building the Everyday App rather than on buying orders. Higher talabat pro adoption, more own-grocery share, faster delivery promise: none of that is a marketing line item. It's operations, product and data.
Which is the part Dubai businesses keep underestimating. E-commerce growth at this scale is not won in a media plan. It's won in catalogue accuracy, inventory sync, order routing, substitution logic, price and availability data flowing between a store's back office and an app in under a second.
What this means if you sell in Dubai
Three practical reads for a UAE business, whether you're a retailer, a distributor, a logistics partner or a software team serving them.
- Your conversion ratio is a diagnostic, not a vanity metric. Talabat's GMV-to-revenue conversion moved from 39% to 40% because its revenue mix changed. Yours moves because of commission structures, delivery fees, returns and subscriptions. If you can't break down your own number that way, you can't tell whether growth is profitable growth.
- Margin compression in the name of growth only works if you can measure the payback. Talabat can, because it reports the investment, the incentive and the outcome separately. Most SMEs here can't, because orders, fulfilment and marketing live in three systems that never talk to each other.
- Own-brand and own-inventory models change your whole data spine. The moment a company moves from taking commission to holding stock, it needs real-time inventory across stores, warehouses and dark kitchens. That's an integration problem before it's an AI problem, and it's the one most Dubai retailers underestimate by about a year.
The infrastructure question nobody puts in the deck
Non-GCC growth of 43% is exciting and also fragile: it depends on currency, payment rails, local supply and delivery partners in markets where the margin for error is thinner. Every one of those dependencies is an integration surface, and every integration surface is where an ambitious e-commerce operation eventually breaks in production.
The businesses that scale cleanly here share a boring trait: they treat their commerce stack as infrastructure. Idempotent order pipelines, catalogues with a single source of truth, observability on fulfilment latency, and an internal API layer that lets them add a new market or a new vertical without rewriting the checkout.
That is exactly the kind of work that doesn't make a press release. It makes the next raise possible.
What I'd check before your next quarter closes
- Can you trace one order from app tap to delivery confirmation without leaving your systems? If not, your growth ceiling is already set.
- Do you know your GMV-to-revenue conversion by vertical, and which line moved it last quarter?
- If 30% of your orders were grocery and retail tomorrow, what breaks first: inventory, cold chain, or customer support volume?
- Are your margins being spent on acquisition, or on capability? Only one of those compounds.
- Have you priced the cost of one bad integration in a new market, before you sign the delivery contract?
One disclaimer worth stating plainly: these are company-reported figures for the quarter ended 30 September 2026, with no independent market-share data in the release, and the non-GCC figure is flattered by currency effects at constant rates.
But directionally, the signal is hard to argue with. Dubai e-commerce growth isn't a story about a market saturating. It's a story about a market professionalising, where the winners are the operators with the cleanest data plumbing underneath the storefront.
If your commerce stack can't survive the next 20% of growth, that's not a marketing problem. It's an engineering one, and it's fixable. That's the work we do at aratech from Dubai Digital Park, and the queue for it is getting longer every quarter.