Iran Skipped the Credit Card. Snapp Pay Just Published What Came Instead.
About 1.3 per cent of the value of what Iranians buy goes on a credit card, against roughly a quarter worldwide, so the country went straight to instalment credit and the instalment lender does the credit scoring itself. Snapp Pay, the largest of them, has now published its own operating and credit data for the twelve months to 22 September 2026: 175 million credit orders, 74,700 merchants, more than 98.5 per cent of instalments ultimately repaid and under 2 per cent unpaid at 90 days. Of 4,170 public facts across 386 public company records in the Tehran Index registry, five carry a repayment or default rate for an Iranian consumer lender, and three of the five arrived with this report.
- Snapp Pay reports that more than 98.5 per cent of its instalments are ultimately repaid, more than 89 per cent of repayments arrive on time or early, and under 2 per cent are unpaid at 90 days, all for the twelve months to 22 September 2026. The figures are company-disclosed and unaudited, and they are the first credit performance numbers any Iranian consumer lender has published at this scale.
- Of 4,170 public facts across 386 public company records in the Tehran Index registry, five carry a repayment or default rate for an Iranian consumer lender. Two are Lendo's, for years ended in 2021 and 2022. The other three arrived with this report. AzkiVam's record lists its default rate among the things it does not disclose.
- Credit cards take about 1.3 per cent of the value of Iranian purchases, online and in store together, against a world average of 20 per cent online and 24 per cent in store on the Worldpay figures the report tabulates. Ninety five per cent of Snapp Pay's users had no bank credit rating when they started with it, so for most of them the lender is the first institution to score them.
- On the company's own four figures, the funnel halves at every step: 45 million eligible, 19.1 million registered (42.4 per cent), 9.5 million who have made at least one purchase (49.7 per cent of registrations) and 4.7 million transacting in the month to 22 September 2026 (49.5 per cent of purchasers). Tehran Index arithmetic on the report's page 45.
- The growth is in shops and outside Tehran. In store four-instalment orders were 8.1 times higher in the month to 22 September 2026 than a year earlier and their value 10.3 times, and 62 per cent of them were placed outside Tehran against 47 per cent a year before. Mature markets run 1 to 4 per cent of in store retail through BNPL; the report puts Iran under 0.5 per cent.
Iran has almost no credit cards. About 1.3 per cent of the value of what Iranians buy, online and in shops together, goes on one. The world average is roughly a quarter. That gap is the single most important fact about Iranian consumer finance, and it is the reason the country’s largest buy now, pay later lender is now running what amounts to a national credit bureau by accident. In autumn 2026 that lender, Snapp Pay, published its first industry report. Most of it is market estimates. Chapter three is not. It is the company’s own operating and credit data, and it is the first time any Iranian consumer lender has put its credit performance in writing at scale.
Why this is the first time
Tehran Index holds 4,170 public facts across 386 public company records. Five of them carry a repayment or default rate for an Iranian consumer lender. Two belong to Lendo, a much smaller platform, and they describe the years to 20 March 2021 and 20 March 2022: a 1.78 per cent default rate by share of customers in the first, and a self-reported 0.5 per cent default rate with 67 per cent of customers paying on time or early in the second. The other three arrived with this report and they are Snapp Pay’s. Everything else in the sector is silence. The record for AzkiVam, one of the larger lenders, lists its loan default rate among the things it does not disclose, alongside revenue, profit and portfolio size. That is the normal state of this industry.
So the interesting part of the Snapp Pay report is not the growth. It is that a lender operating at this scale decided to publish the thing lenders do not publish. The company frames it that way itself: it calls the three measures its answer to what it describes as the most important regulatory question facing BNPL globally, whether uncollateralised credit at large scale turns into deferred debt. Its answer is no. It is the company’s answer about its own book, unaudited, and it should be read as that. It is still the only answer anyone in Iran has shown their work for.
The card that never arrived
The report opens with the gap, and the gap explains the product. Iranian payment rails are excellent and almost entirely debit. Card to card transfer is universal. What does not exist is revolving consumer credit at the point of sale, because the cost of money to a lender in Iran is high and the credit scoring layer that a card issuer needs was never built. So the country did not pass through a credit card era on its way to instalment credit. It went straight to instalment credit, and the instalment lender does the scoring itself.
| Region | Online retail | In store |
|---|---|---|
| World | 20% | 24% |
| North America | 33% | 41% |
| Latin America | 31% | 25% |
| Europe | 19% | 18% |
| Middle East and Africa | 18% | 11% |
| Asia Pacific | 11% | 13% |
| Iran | about 1.3%, both channels together | |
The consequence shows up in who Snapp Pay lends to. Ninety five per cent of its users had no bank credit rating at all when they started with it, and 41 per cent are under 30. The company is not competing with a credit bureau. In most of these cases it is the first institution to form a view on whether the person repays.
What came instead, on the company’s own estimate
| Period | Thousand bn toman | Share of online retail |
|---|---|---|
| Year to 20 March 2023 | 4 | 4% |
| Year to 19 March 2024 | 16 | 7% |
| Year to 20 March 2025 | 58 | 14% |
| Year to 20 March 2026 | 132 | 17% |
| Year to 20 March 2027, forecast | 400 | 28% |
Two things in that table are worth separating. The toman column is nominal in years of serious inflation, so the ten times move from 4 to 132 is not a ten times move in real purchasing. The share column is inflation neutral, and it is the one that carries the argument: BNPL went from 4 to 17 per cent of Iranian online retail in four years. On the same page Snapp Pay puts the global average at 12 per cent, and puts BNPL at under 2 per cent of all Iranian retail, because online retail itself is only about 7 per cent of what the country buys. Our own read of the national retail measurement landed on a similar order for the online share, from an independent source.
Against the countries the report tabulates, Iran’s 17 per cent online BNPL share sits level with Australia and above the United Kingdom at 12 and the United States at 8. The in store column is the opposite story. Mature markets run 1 to 4 per cent of in store retail through BNPL. Iran runs under 0.5. That is the number Snapp Pay is spending against, and the growth figures show it: in the month to 22 September 2026, against the same month a year earlier, in store four instalment orders were 8.1 times higher and their value 10.3 times. Sixty two per cent of those orders were placed outside Tehran, against 47 per cent a year before.
The funnel, and what it implies
Page 45 prints the whole thing as one stack, which is unusual and useful. Snapp Pay says 45 million Iranians are eligible for its credit. 19.1 million have registered. 9.5 million have made at least one successful purchase. 4.7 million transacted in the month to 22 September 2026. On those four figures, registration converts 42.4 per cent of the eligible pool, a first purchase converts 49.7 per cent of registrations, and monthly activity runs at 49.5 per cent of everyone who has ever bought. Half, then half, then half.
Two derivations the report does not make. It says 175 million successful credit orders in the twelve months and an average of 24 orders per user over the same year; those two together imply about 7.3 million users transacting across the year, out of the 9.5 million who have ever bought. And 175 million orders over twelve months averages roughly 479,000 a day, against the more than one million a day the company reports for the final month of that same year. The business more than doubled inside the period it is describing, which is consistent with the 136 per cent order growth it reports for that month.
Against the lenders that do file accounts
| Lender | Period | Volume | Users | Merchants |
|---|---|---|---|---|
| Klarna | Year to 31 Dec 2025 | $127.9bn GMV | 118m active consumers | 966,000 |
| Affirm | Year to 30 Jun 2026 | $50.2bn GMV | 27.8m active consumers | 571,000 active |
| Snapp Pay | 12 months to 22 Sep 2026 | 175m credit orders, no GMV published | 4.7m monthly transacting, 9.5m have ever bought | 74,700 unique |
The comparison stops earlier than it looks. Klarna’s credit figure is a provision for credit losses running at 0.65 per cent of GMV in the fourth quarter of 2025. Snapp Pay’s is a 90 day non-repayment rate on instalments. Those are different measures of different things and ranking one against the other would be wrong. Snapp Pay publishes no GMV in toman and no loan book, so the only volume measure that crosses is the order count. On merchants, Snapp Pay’s 74,700 is about 13 per cent of Affirm’s 571,000, in one country against Affirm’s market. We did not read company or regulator figures for Afterpay, Tabby or Tamara in this run, so they are absent rather than small. The report itself gives only their founding years.
What only this desk holds
The same company, measured twice, six months apart. Snapp Group’s own annual report put Snapp Pay at 13,548,854 registered accounts and 22,165 stores for the year to 20 March 2026. The industry report, about six months later, says 19.1 million registered users. Registered users rose 41.0 per cent in half a year. The two merchant figures are not like for like, one counting stores and the other unique merchants including social commerce sellers, and they are not compared here.
The registry. The operating company behind the brand, Atlas Tejarat Novin Diba, sat at 1,000,000 rial of registered capital from its incorporation in February 2015 until December 2021. It then rose in steps to 2 trillion rial on 10 January 2024 and 6 trillion rial on 30 September 2025, both paid in by its parent Ide Negaran Atlas Roomak. Registered capital is a nominal legal figure and is not a valuation of anything. What it does show is when the parent decided this was a balance sheet business rather than a feature.
The App Index. Snapp Pay has no standalone app in any of the desk’s store captures. It runs inside the Snapp super app, which went from 55 million to 59 million installs on Cafe Bazaar between 29 June and 2 November 2026. Its nearest group-backed rival, Digipay, went from 5.8 million to 6.9 million over the same weeks, and Digikala’s annual report puts Digipay credit purchases up 221 per cent in the year to 20 March 2026 across 5,351 accepting brands and stores. Distribution is the asymmetry: one of these lenders has to be downloaded, and the other is already on the phone.
What we do not know
No GMV and no loan book in toman from Snapp Pay, and no audited statement of any kind, so every company figure here is the company’s own claim about itself. No breakdown of the 98.5 per cent: we do not know what share of the shortfall is written off against what share is still being collected, nor the loss given default. No cost of funds, which in Iran is the number that decides whether a lending book earns anything. The 45 million eligible figure has no stated method. The report’s own search page contradicts itself, giving 38 and 42 million monthly referrals in two places, so nothing from it is used here. The jobs and consumer benefit figures elsewhere in the report are modelled estimates and are not used. No dollar figure is printed for any Iranian period, which is a standing rule of this desk and it bites here: the toman growth cannot be read as real growth without a deflator we are not going to improvise.
This piece is about disclosure, not compliance. We have not read the rulebook that governs instalment credit in Iran and make no claim about anyone’s standing under it.
The earlier benchmarking read, which sized the market from industry estimates and set Iran against Gulf platforms before any of this first-party data existed, is Iran’s BNPL Market, in Global Context. The structural reason the sector exists at all is in The 2% Problem in Iran’s Lending Market. Snapp Pay’s sourced figures, with their periods and evidence, sit on the Snapp Pay record.
Cite as: Tehran Index, “Iran Skipped the Credit Card. Snapp Pay Just Published What Came Instead.”, 3 November 2026. Operating and credit figures from Snapp Pay, Iran BNPL industry report, autumn 2026, pages 14, 17, 19, 27, 44, 45, 46, 47, 49, 50, 51, 53 and 57, read as page images. Klarna from its full year 2025 results of 26 February 2026 and its fourth quarter 2025 earnings release. Affirm from its fiscal fourth quarter 2026 earnings supplement of 27 August 2026 and its Form 10-K for the year to 30 June 2026. Registry and App Index figures computed from the Tehran Index record on 3 November 2026. Conversion rates through the funnel, the implied transacting-user count, the daily order average and the change in registered users are Tehran Index derivations. tehranindex.com
Frequently asked
Snapp Pay reports that under 2 per cent of its instalments are unpaid at 90 days and that more than 98.5 per cent are ultimately repaid, both for the twelve months to 22 September 2026, in its own industry report published in autumn 2026. The figures are company-disclosed and unaudited, the company does not break down what share of the shortfall is written off against still being collected, and it publishes no loss given default. It is not the same measure as the provision for credit losses that Klarna and Affirm charge against volume, and the two should not be ranked against each other.
In the twelve months to 22 September 2026 Snapp Pay reports 175 million successful credit orders across 74,700 unique merchants in 804 cities, with more than one million credit orders a day on average in the final month of that period and a record of 82,000 orders in one hour. It reports 19.1 million registered users, 9.5 million who have made at least one successful purchase and 4.7 million transacting in the month to 22 September 2026. It publishes no transaction value and no loan book.
Credit cards take about 1.3 per cent of the value of Iranian purchases, against roughly a quarter worldwide. Iranian payment rails are strong and almost entirely debit; what was never built is revolving consumer credit at the point of sale, because the cost of money to a lender is high and the credit scoring layer a card issuer needs does not exist. Instalment lenders filled the gap and do the scoring themselves: Snapp Pay says 95 per cent of its users had no bank credit rating when they started.
Snapp Pay estimates it at 132 thousand billion toman for the year to 20 March 2026, up from 4, 16 and 58 in the three years before, and forecasts 400 for the year to 20 March 2027. On the same estimate BNPL is 17 per cent of Iranian online retail, against a global average the report puts at 12 per cent, and under 2 per cent of all Iranian retail, because online retail is only about 7 per cent of what the country buys. These are the company's analysis of public statistics rather than reported totals, and the toman figures are nominal in years of high inflation, so the growth is not real growth.
Klarna reported $127.9 billion of GMV, 118 million active consumers and 966,000 merchants for the year to 31 December 2025. Affirm reported $50.2 billion of GMV, 27.8 million active consumers and 571,000 active merchants for the year to 30 June 2026. Snapp Pay reports 175 million credit orders and 74,700 merchants for the twelve months to 22 September 2026 but publishes no transaction value, so volume can only be compared in orders. Klarna and Affirm are audited public filers; nothing in the Snapp Pay report is audited.
This piece is built from the same registry the platform runs on: company records with sourced, period-stamped facts and a link on every claim. Access opens to a limited first cohort. Registering costs nothing, commits you to nothing, and gets a reply from the research desk within two business days.
Register for accessReader discussion
Create an account to comment. Members also get a watchlist and the Friday digest.