Iran’s Biggest App Store Published the Money for Five Years. Then It Stopped.

Cafe Bazaar handles most of Iran's Android app distribution, and for five straight years it published what developers earned through it. The last figure is 280.5 billion toman, for the year to 20 March 2021, from 27.1 million purchases by 3.73 million buyers. Its own five-year series shows the payout rising 4.14 times in nominal toman while the number of people who paid for anything rose 11 percent. Then the money left the report: the two later annual editions Tehran Index holds, one of ten pages and one of 110, carry no figure in toman or rial anywhere.

For five straight years Iran’s dominant Android app store told the public exactly how much money passed through it. Every year Cafe Bazaar published an annual report, in Persian, carrying one line almost no company in Iran publishes: what the developers on its platform earned. In the year to 20 March 2021 that figure was 280.5 billion toman, from 27.1 million purchases made by 3.73 million people.

That is the last such number anyone has. The store has published two more annual reports since, one of ten pages and one of a hundred and ten, and neither carries a figure in toman or rial anywhere in it.

What the five years actually show

The 2020-21 edition is the useful one, because it does not just report its own year. It prints six five-year series on three pages, each back to 2016-17, which makes it the only document that measures the Iranian app economy consistently over time from inside the platform that runs most of it.

Read down the first column and the story looks like growth: what developers earned rose 4.14 times in five years. Read across to the fourth column and it is a different story. The number of people who paid for anything on the store rose from 3.36 million to 3.73 million. That is 11 percent in five years, in a country whose smartphone base was expanding the whole time, and the same store’s registered accounts went up 2.68 times over the same period.

So the money did not come from more buyers. It came from the same small group buying more often, and from prices. Transactions per buyer, which the store computes itself, went from 3.87 a year to 7.26, a rise of 1.88 times. Divide the payout by the buyers and the average developer revenue per paying user went from about 20,200 toman a year to about 75,200, a rise of 3.73 times. Those last two figures are ours, computed from the two series printed on the same page.

The denominator is the part worth holding on to. In 2016-17 the year’s buyers were 17.7 percent of all accounts ever registered on Bazaar. In 2020-21 they were 7.3 percent. That is a flow measured against a stock, so it is a rough gauge rather than a conversion rate, but the direction is not ambiguous: the store added users much faster than it added payers.

Why the 4.14 times is not a growth rate

Everything above is in toman as the company printed it, and that is a deliberate limit rather than an oversight. The five years from 2016-17 to 2020-21 covered one of the sharpest falls in the rial’s value in modern Iranian history. The report prints no dollar figure, no constant-currency figure and no inflation adjustment, and Tehran Index does not supply one here, because no dated period rate for each of those five years sits on our record and a rate invented for the purpose would be worse than the gap.

What that means in practice: 4.14 times nominal over five years is not evidence that the Iranian app economy grew 4.14 times. It is evidence that the number of toman changing hands rose that much. The unit-count series beside it, the ones that do not depend on a currency, are the ones to argue from, and they say something much more sober. This is the same measurement problem we set out in Iran’s Exit Problem Is an Exchange Rate Problem, seen from the operating side rather than the investor’s.

Then the money disappeared from the report

The two later editions the desk holds are not shorter or thinner. The 2024-25 one is the longest edition the desk holds, at 110 pages against 17 for the first one in this series. They simply count different things.

Look at what the newest edition does measure and a pattern comes out of it. More than 166 million phones registered on Bazaar since launch, more than 28 million of them in the year. 1,017 million app and game updates, up 16 percent. 27,400 active developers. 67 thousand publication requests and about four thousand refusals. 170 thousand incoming user support calls, down 24 percent. And, in place of the payout line, more than 51.5 million installs delivered through the store’s own advertising tools, from more than 7,400 advertising packages.

That is a reach document. It describes an audience and the volume of traffic that can be sold against it. The earlier reports described a marketplace and what it paid out. Both are honest accounts of a business. They are accounts of two different businesses.

Two things changed around the same time

We state the sequence and stop there, because a sequence is what the record supports.

In January 2025 Cafe Bazaar was acquired by Tapsell, which our registry describes as the dominant mobile performance and digital advertising network in Iran. The acquisition is not a rumour: it is stated in the audited annual report of a foreign investor that held an indirect economic interest in the store, which is how it reached our record at all. Both companies now sit under the same holding company. The buyer’s chief executive put the price at about 2,000 billion toman on the record.

And the 2024-25 report tells you who wrote it. Its colophon, on page 108, dates publication to the Iranian month that ended on 22 September 2025, and says the report was prepared and compiled by Bazaar in cooperation with ILIA Management Consulting, with infographics and layout by The Right Agency. The file’s own embedded document properties say the same thing in English, adding that the document was commissioned by Bazaar and that the analysis and content were prepared by the consultancy using Bazaar data. The file was created on 23 September 2025.

The same report records one more thing worth noting beside the rest: admission to the Farabourse, obtained, it says, after four years of effort, and it calls Bazaar the second active digital economy platform to get it.

What is left to measure from outside

When a company stops publishing a number, the number does not become unknowable. It becomes somebody else’s job.

The Tehran Index App Index now reads the Cafe Bazaar store directly. On 16 capture days between 29 June and 5 October 2026 it recorded 753 listing-day rows across 69 Android packages on that store, each carrying an install count and a rating. Not one carries a price, a payout or a revenue field, because the store does not publish one on a listing page. We can measure how many people install an Iranian app. Nobody outside the company can now measure what the developer was paid for it.

That is the gap this piece exists to mark. Across the whole Tehran Index record, Cafe Bazaar is the second most documented company we hold, 257 public facts against Digikala’s 307, out of 3,703 public facts on 385 public company records on the morning of 9 October 2026. Fifteen of those 257 carry a money figure. Two of the fifteen come from outside the company, the registered capital on its registry file and the price the buyer stated. The other thirteen all come from Cafe Bazaar’s own reports, and every one of the thirteen belongs to a period that ended on or before 20 March 2021.

A different way to put the same thing: the best-documented app store in Iran has told the public nothing about money for more than five years, and we only know that because we read every edition it has published since.

What we do not know

We do not hold an annual edition for the year to 20 March 2022 or for the year to 19 March 2024, and we have not established whether either was published. The claim here is about the six editions we hold and about the last payout figure anyone has found, not about editions that may exist unseen. We do not know whether Bazaar reports payouts privately to its own developers; a developer console is not a public document. We do not know why the disclosure stopped, and nothing above should be read as saying the ownership change caused it. We do not know Cafe Bazaar’s own revenue in any year: what it published was always what developers earned, never what it kept.

Two measurement cautions belong here too. The store changed its headline audience measure between these editions, from active installs, which it defines as devices carrying the app in the last two months, to monthly active users, which first appears in the 2024-25 report at 50 million; those are different quantities and we do not join them into one line. And on page 86 of that same report, new apps published read 74,185 for the year to 19 March 2024 and 17,706 for the year to 20 March 2025, with games falling from 46,117 to 18,757 beside them. We print both and draw no trend from them. A fall of roughly 70 percent in one year, on one page, with no change of definition stated, is more likely to be a change of definition than an event, and until we can establish which, it is not a finding. It is a question, and it is now on our list.

For how the store’s present-day numbers compare with its nearest competitor’s, and why the two disagree about every app they both carry, see Iran’s Two App Stores Disagree About Every App.

Cite as: Tehran Index, “Iran’s Biggest App Store Published the Money for Five Years. Then It Stopped.”, 9 October 2026. Five-year series from Cafe Bazaar’s own annual report for the year to 20 March 2021, pages 8, 9 and 11, read as page images. Later editions from the company’s annual reports for the years to 20 March 2023 and 20 March 2025, read in full. Fact counts and App Index coverage computed from the Tehran Index registry, facts layer and App Index on 9 October 2026. tehranindex.com