Ninja's 3 Year Lesson: Speed Proves the Operator, It Never Proves the Model

Three years to a unicorn, and the decade that came first

On 1 July 2025 a Saudi company that did not exist four years earlier raised about 250 million dollars at a valuation of 1.5 billion. Riyad Capital led the round. The company was Ninja, founded in 2022, and the headline wrote itself: the fastest unicorn in the Kingdom's history, three years from nothing to a billion and a half.

The three years is the part everyone repeats. It is the hook in the plan behind this series, it is the first line of most of the coverage, and it is true.

It is also the wrong unit. The company is three years old. The operator is not. Ninja was founded by Ebrahim Al-Jassim, who founded HungerStation, the app that taught Saudi Arabia to order food on a phone. He ran it for roughly a decade. In April 2019 Delivery Hero removed him and the entire executive team in a single day, and he spent four years litigating it, including a memorandum opinion from the United States District Court for the District of Columbia in May 2021. Delivery Hero bought out the remaining 37 per cent of HungerStation in July 2023 for 297 million dollars.

So the three-year sprint is the second half of something much longer. The playbook, the supplier relationships, the operating instincts and the investor access all predate the company by about ten years, and none of that shows up in the founding date.

There is a detail that makes the point better than any argument. In June 2026 Ninja was reported to be weighing a bid for Delivery Hero's Middle East assets, with HungerStation first on the list. The three-year-old unicorn is trying to buy the eleven-year-old company its own founder started and was fired from. That is not a story about speed. It is a story about a man finishing something.

What was already built before Ninja opened a single store

Every case study in this series has a section like this, because the famous number is almost always sitting on top of infrastructure somebody else paid for. Ninja's is unusually well documented.

The payment rails were finished. Saudi Arabia runs near universal mada debit coverage and electronic payments crossed 85 per cent of retail transactions. A company launching in 2022 did not have to teach anyone to pay with a phone, which is most of what a delivery business spends its first years doing in a less developed market.

The habit was finished too, and Al-Jassim had helped finish it. HungerStation, Jahez and Mrsool spent the decade from roughly 2012 onward training Saudi consumers to expect a stranger at the door in under an hour. By 2022 the behaviour was ordinary. Ninja did not create a market, it entered a mature one with a faster promise.

The regulator had already formalised the sector. The Transport General Authority licenses delivery operators and publishes quarterly order counts for the whole Kingdom, which means the category had a compliance framework and a public scoreboard before Ninja existed.

And the geography was favourable in a way it simply is not in the markets where this model died. Riyadh is dense, hot, and car dependent in a manner that makes a fifteen minute grocery run genuinely unattractive. The Transport General Authority's own bulletin puts Riyadh at 45.04 per cent of all national delivery orders, with Makkah at 21.17 per cent and the Eastern Province at 15.87 per cent. Three regions carry more than four fifths of the demand, which is about as concentrated as a dark store operator could ask for.

None of this makes what Ninja did easy. It makes it possible, which is a different claim, and it is the claim the three-year headline quietly drops.

The decision, which was operational and not a campaign

The move that defines Ninja is not a piece of marketing. It is a balance sheet choice.

HungerStation and Jahez are marketplaces. A restaurant cooks, a driver collects, the platform takes a commission and never owns the food. It is asset light, the gross margin is a fee, and the hard problem is supply of restaurants.

Ninja holds stock. It runs its own dark stores, buys inventory, employs the picking, and sells the goods. Roughly 100 dark stores across 28 cities, with a stated plan to pass 200, delivering groceries and daily essentials in 25 to 30 minutes. That makes it a retailer with a logistics network attached, not a marketplace with an app attached.

This matters for reading every number that follows. When a marketplace reports revenue, it is reporting commission. When Ninja reports revenue, it is reporting the full value of the basket, because the basket was its own. Two companies in the same sector with the same revenue line can be running businesses an order of magnitude apart in size, and quick commerce sits on the flattering side of that distinction.

It also explains why the category kills people. Owning stock means wastage, rent, picking labour and a delivery cost that does not fall with scale the way software costs do. The model works when enough orders per store per day come out of a small enough radius, and it fails when they do not. That single ratio is the whole business, and no company in the category publishes it.

What a 1.5 billion dollar valuation actually records

The round was announced on 1 July 2025: about 250 million dollars, reported by some outlets as 254 million, led by Riyad Capital, described as pre-IPO, at a post-money valuation of 1.5 billion dollars.

Start with what that sentence is. A valuation is a price somebody paid for shares on a particular day. It is real evidence that sophisticated investors looked at the accounts and were willing to transact. It is not revenue, not profit, not a customer, and not a marketing result.

Then note the size of it relative to everything before. Ninja's total lifetime funding is reported at roughly 284 million dollars. The July 2025 round is therefore about 88 per cent of every dollar ever put into the company. Almost the entire capital base arrived in one cheque, fourteen months ago, which makes the valuation less a trajectory than a single data point.

The company's own figures alongside it: about 1 billion dollars of revenue in 2025, a target of 1.6 billion for 2026, and a claim to be profitable in its core operations. All three are company stated and none is audited.

And then the thing that has not happened. In the fourteen months since that round, no primary round has repriced the company. The 1.5 billion is still quoted in the present tense in September 2026, exactly as Tamara's billion was in Episode 16, and the intended repricing event keeps receding. Banks were reported mandated by September 2025. Bloomberg reported the company gauging appetite in March 2026. By May 2026 the listing had moved to late 2026 or early 2027, still preliminary, with a private raise named as an alternative. Citigroup, Goldman Sachs, UBS and Riyad Capital are on the mandate for a raise of about 1 billion dollars.

One structural detail is worth stating without alleging anything by it: Riyad Capital led the private round that set the 1.5 billion price and is also one of the four banks mandated on the offering that would test it. That is ordinary in Gulf capital markets, and it is still a reason to treat the private mark as a negotiated number rather than an independent one.

The scoreboard

Every figure in this table is a price, a capacity, a company statement or a regulator count, and the table labels each one. The 1.5 billion dollar valuation is a price paid for shares on 1 July 2025 and has not been retested by a later primary round. The revenue, the 2026 target, the store count and the profitability claim are the company's own statements and are not audited; in quick commerce the company owns the inventory, so a revenue figure is the full value of goods sold rather than a commission, which makes it much larger than a marketplace's revenue for the same economic activity. The implied multiple is arithmetic on those two company stated numbers and is offered as a scale check, not as a valuation. The Jahez figures are audited and taken from its published results, and its revenue base is not directly comparable to Ninja's for the reason just given; riyal figures convert at 3.75 to the dollar. The delivery order counts and the regional split are the Transport General Authority's own quarterly bulletins for the entire Kingdom across all providers, not Ninja's orders. The Getir valuations are prices set on dates by rounds that are now historical, and the 374 million dollar figure is the company's own asset analysis submitted in litigation. Several publishers carrying these figures are blocked by the egress proxy used to research this piece, including forbes.com, techcrunch.com, stockanalysis.com and companiesmarketcap.com, so those numbers were confirmed through search results and through outlets quoting them rather than by opening the pages directly.

The autopsy: a great operator, and a scoreboard that cannot prove it

Take the numbers in order and ask what each one is.

The 1.5 billion dollars is a price set on 1 July 2025 and not retested since. Episode 16 showed Tabby's headline move three times in eleven months, from 3.3 to 4.5 to 6.5 billion, and warned that a figure quoted in the present tense is usually a figure with a date hidden inside it. Ninja is the opposite failure mode of the same mistake: the price has not moved at all in fourteen months, and a mark nobody has refreshed is as unreliable as one that keeps changing, in the other direction.

The 1 billion dollars of 2025 revenue is company stated, unaudited, and gross. Ninja owns the inventory, so the revenue line is the full basket value rather than a commission. That is not a criticism, it is an accounting fact, and it changes what the valuation means: 1.5 billion over roughly 1 billion of gross revenue is about 1.5 times sales, which is a grocery and logistics multiple. The unicorn headline is a function of how much merchandise moved through the company, not of a premium the market placed on its brand.

The comparison that settles this is public, audited and in the same country. Jahez has been listed on the Tadawul since 2022. Its 2025 revenue was 2,323.6 million riyals, up 4.7 per cent, and its net profit fell 61.1 per cent to 73 million riyals from 188 million in 2024, with its chief executive attributing part of the fourth quarter to 55 million riyals of one-off items. Its market capitalisation has sat around 2.6 to 2.8 billion riyals in 2026, roughly 700 to 760 million dollars, on revenue of about 620 million dollars. So the only Saudi delivery company that has to publish audited accounts saw its profit collapse by three fifths in the same year Ninja was crowned, and the public market prices it at a little over one times revenue. The private mark on Ninja is in the same neighbourhood. Nothing in either number is a marketing result.

Now the profitability claim, because the load-bearing word is core. A segment-level profitability statement from a private company is the company's own, unaudited, and defined by the company. Core places costs outside itself, and Ninja has obvious candidates: three new country markets in Bahrain, Qatar and Kuwait opened inside roughly twelve months, a build from about 100 dark stores toward more than 200, and head office. A company can be profitable in its core operations and lose money as a company in the same year, and both sentences can be printed truthfully in the same press release.

The growth deserves better than scepticism, so here it is plainly. The Transport General Authority publishes the market-wide order count every quarter: 79.6 million orders in the first quarter of 2025, more than 101 million in the second, and 132 million in the second quarter of 2026, up 30.5 per cent year on year across every provider in the Kingdom. A sector growing about 30 per cent a year is a wave. Ninja targeting 1.6 billion against about 1 billion is roughly 60 per cent, which is about twice the market, and outgrowing your category by that margin at that size is genuine operating performance. Episode 6 was about a company that mistook a wave for its own strength. Ninja is not that company. It is swimming hard, and it is also swimming with the current, and the honest reading holds both.

Which leaves the number that would actually settle it, and it is missing. Orders per dark store per day, and the fully loaded contribution margin per order after delivery, picking, wastage and rent. That ratio is the entire business, it exists inside Ninja to a decimal place, it is what killed Getir in Europe and what put Nana in front of a Riyadh court, and it has never appeared in a single announcement. Every Stage 3 case has a number like this. On this one it is not a detail, it is the thesis.

The control group, which this category has in embarrassing quantity

Quick commerce is the most copied and most comprehensively failed model of the last decade, which means the counterfactual is free to look up. Episode 12 made the general point: when a tactic has been widely copied, the copies are the control group.

Getir is the same model run with far more money. Mubadala backed it to a reported 11.8 billion dollar valuation in 2022. That fell to about 2.5 billion, then to roughly 1 billion. In April 2024 it withdrew from the United Kingdom, Germany, the Netherlands and the United States to concentrate on Turkey, stating that the abandoned markets produced only 7 per cent of its revenue. In an analysis submitted to a Dutch court it valued its group assets at 374 million dollars, and its founders are suing Mubadala for at least 700 million. Gorillas, which Getir had bought in December 2022 for about 1.1 billion euros, ceased operations entirely in May 2024. Jokr pulled out of the United States. Buyk and Fridge No More closed. Zapp cut staff.

Here is where an honest autopsy has to change direction, because the European graveyard does not transfer automatically and pretending it does would be the easy contrarian move this series exists to avoid. The GCC is structurally different in the two ways that decide dark store economics. Urban density in Riyadh, Dubai and Kuwait makes last-mile cost work where sprawl does not, and willingness to pay for convenience is high enough to reduce the dependence on permanent discounting that hollowed out the European players. Redseer, which covers the category closely, calls the GCC one of the most credible environments globally for sustainable dark store economics and projects quick grocery reaching 89 per cent of online grocery and around 14 per cent of all grocery retail in Saudi Arabia and the Emirates by 2030. The plan behind this series says speed of execution in quick commerce can mint a unicorn in three years. On the evidence, that is probably right, and it should be said plainly rather than argued away.

So the useful control group is not European. It is domestic, recent, and it ran in the same market during the same years Ninja was winning. Nana raised about 208 million dollars across five rounds from backers including Kingdom Holding and STV, and the Commercial Court in Riyadh opened financial reorganisation proceedings for its operating entity, giving creditors 90 days to file claims. Shgardi buckled in late 2025. Careem retreated from Saudi grocery about thirteen months after entering it. Rabbit, which arrived from Cairo promising fifteen minute delivery, folded its Saudi business and went home. Same country, same category, same window, same favourable density. Four casualties.

And then the finding that narrows the plan's angle more than anything else in this episode. Redseer's own account of who has led quick retail in Saudi Arabia runs: Omnichannels, then Nana, then HungerStation, then Ninja. The number one position in this category has changed hands four times. Nana did not merely fail, it failed after leading. Whatever Ninja is holding, the category's history says it is a lease rather than a freehold.

The plan's BMD angle is that operational excellence is itself a brand story. Half of that survives and half of it does not. The operational excellence is real and it is the reason Ninja is where it is. Calling it a brand story is the part to be careful with, because a brand is what remains when the operation is briefly beaten on price or speed, and in a category that has replaced its leader four times in a decade nobody has yet demonstrated that the operation leaves anything behind when it slips. Ninja's own move on HungerStation suggests the company knows this: you do not buy an incumbent's brand and customer base if you believe your operating advantage alone is durable.

The right order: the unit, then the density, then the brand

The portable version of this, for a company that will never raise 250 million dollars, is an order of operations. Quick commerce makes it unusually legible because the category has run the experiment to destruction on four continents.

First, close the unit before you buy the demand. The question is not whether people want fast delivery, because they always do at a subsidised price. It is what one order contributes after the driver, the picker, the wastage and the rent, and how many orders a single location does per day. Getir raised billions without closing that gap and the gap eventually closed the company. If the unit does not work at 40 orders a day per store, marketing that takes you to 60 buys you a faster loss.

Second, buy density rather than awareness. Dark store economics are a function of orders inside a delivery radius, which means the correct growth spend is geographically concentrated until a location is saturated, and then repeated somewhere else. A national awareness campaign against 100 stores spreads demand across a surface the network cannot serve profitably. Ninja's 28 cities against 100 stores is roughly three and a half stores per city, which is a density strategy rather than a coverage one.

Third, and only then, spend on the brand, on the specific grounds that the operation will eventually be matched. The reason to build a brand in an operationally led category is not that it sells the first order, because speed and price do that. It is that it holds the customer through the quarter when a competitor is faster or cheaper. Four changes of leadership in Saudi quick retail is the argument for this, not against it.

At Saudi mid-market scale the numbers shrink and the sequence does not. The demand data is free and public: the Transport General Authority publishes the Kingdom's quarterly order count and its regional split, so anyone can know before signing a lease that Riyadh is about 45 per cent of national delivery orders, Makkah about 21 per cent and the Eastern Province about 16 per cent. That is a targeting decision available at no cost to a company with four vans, and most companies in the category still make it by intuition. Work out the contribution per order on a spreadsheet, pick one district rather than one country, and hold the brand spend until the unit closes. The sequence is the strategy.

The takeaways

Three years measures the company, not the operator. Ninja's founder spent roughly a decade building HungerStation, was removed from it in 2019, litigated for four years, and is now reported to be bidding for it. You cannot raise the decade.

A valuation is a price on a date, and a price nobody has retested is as unreliable as one that keeps moving. Ninja's 1.5 billion was set on 1 July 2025 and is still quoted in the present tense fourteen months later, while the listing that would test it has slipped twice.

Check what a revenue line contains. Quick commerce owns its inventory, so the revenue is the full basket rather than a commission, and 1.5 billion over about 1 billion of gross revenue is a grocery multiple wearing a technology headline.

Read the qualifier. Profitable in its core operations is a company stated, unaudited, company defined sentence, and the word core is where three new country markets and a doubling of the store estate can be parked.

Say plainly when the performance is real. A sector growing about 30 per cent a year is a wave, and growing at roughly 60 per cent on top of it is genuine outperformance. Both halves of that belong in the same paragraph.

The category has replaced its leader four times: Omnichannels, Nana, HungerStation, Ninja. Nana raised about 208 million dollars, led the market, and ended up in court supervised reorganisation. Operational leadership is a lease.

Frequently asked questions

This article is part of BMD's marketing case-study series. Episode 17 is the second of the Gulf run. Episode 18 is the largest technology exit the Middle East has produced, a company sold for 3.1 billion dollars to the competitor it had spent years holding off in its own region, with the price split between cash and notes and the super-app later sold on again for a fraction of it. That is Careem.

Sources and further reading

About BMD

Most companies don't have a marketing problem. They have a marketing department that was never built. BMD is a boutique consultancy that installs structured, measurable marketing departments inside mid-market companies across the GCC. We don't run your campaigns, and we don't hand you a strategy deck and leave. We build the operating system: the structure, the measurement, and the ownership that turn marketing into a function leadership can rely on. The method is the BUILD framework, published and practiced: a book, an online program, a community of Gulf founders and marketers applying it, and diagnostics that replace assumptions with measurement. Delivered in Arabic and English, founder-led.

Redha Alayesh

A marketer with a software engineer's discipline and a scientist's mindset. Across 40+ marketing departments in the GCC, he built the BUILD framework to solve the problem he kept finding: capable marketers trapped inside companies that never built them a department.