Institutional research on Iran’s private technology economy.

Company filings, corporate registry records, official gazette notices, exchange disclosures, company channels and trade press, read in the original language. Every fact carries four things: the value as the source stated it, the period it covers, how strongly we hold it, and a link to where it came from. 623 facts across 370 company records today, each one traceable.

Not the language. Disambiguation. Respina and Raspina are different companies. Products file as companies. A parent company record routinely hides an empty subsidiary underneath it. Founders register in UAE free zones while operating from Tehran, so a foreign address on the paperwork is not a foreign business, and treating it as one misleads exactly the reader the record exists for. Getting those calls right, and recording each one so it is never argued twice, is the work.

Sources are read daily and signals are linked to the companies they concern on the way in. Every record carries the date it was last reviewed, and the coverage report states what is thin. We publish the age of a fact rather than implying everything is fresh.

We hold both and say so. A figure that diverges materially between credible sources is published with both anchors and the divergence named, never averaged into one clean number. Where a conversion is involved the rate and its date are stated, because a currency figure without its rate is not a figure. The full rules are published at /standards.

Valuations of private companies, because they cannot be verified from the outside and a record that guesses at them is worth less than one that declines. Registered capital is a legal filing, not a valuation, and we never present it as one. Registered accounts are not users and driver accounts are not drivers, so we carry the active figure alongside the registered one wherever a source gives both. The perimeter is business, technology and markets.

It gets corrected in public, at /corrections, with what changed and why. Every record carries its evidence state so you can see what we hold and how strongly before you rely on it, and a live audit manifest states what is in production at any moment. A desk that never publishes a correction is not being careful, it is not checking.

370 company records across 23 sectors, with 12 at full dossier depth and the rest at varying completeness. The gaps are published rather than hidden: the coverage report names which sectors are thin and which fields are missing. A record that knows its own blind spots is more useful than one that pretends not to have any.

Internal use across your organisation, and citation with attribution. No redistribution or resale. The full terms are published at /data-terms.

A reply within two business days from the research desk, and one question: which sectors and companies matter to you. That answer shapes what gets deepened next. The first cohort is contacted before access opens, and terms are agreed in writing beforehand.

Access at this level means a dossier scoped to your questions, and research requests answered by the people who built the record. One desk can hold that standard for a small number of organisations. A longer list would mean a worse one.

We are not naming a date we cannot hold. There is no published price list; terms are agreed directly with each organisation at scoping, in writing, before anything begins, and the first cohort holds its terms before commercial pricing is introduced. Everything published now stays free to read. Registering costs nothing and commits you to nothing.