Context
Iran's used-phone market ran on classified listings with no way to know what a phone was actually worth or whether it worked. Tank added inspection and pricing to that transaction, the equivalent of what Karnameh does for used cars, narrowed to mobile.
The reframe
Our instinct was to buy search ads against terms like 'used iPhone'. But when I looked at what people were actually typing, almost none of it was that. It was a specific model and a specific worry: is this price fair, is 82% battery health bad, is this the real model. Thousands of those, every month, asked at the moment of maximum doubt. The market was not looking for a marketplace; it was asking a question no marketplace answered. So instead of paying to interrupt those questions with an ad, we built the answer: a page per model, backed by real inspection and pricing data, that met the doubt head-on. The traffic followed because we were the only ones replying.
What I built
I owned the product and the design: the inspection flow a seller walked through, the pricing that turned a raw phone into a number both sides could trust, and the model pages that turned that data into something search could find. The through-line was making a private, anxious transaction legible enough to happen between two strangers. Beyond the product, I drove the deals that gave it real supply and real trust. I brought the most modern phone-and-accessories mall in Qom online, turning every shop in it into a single digital marketplace. I partnered with the province's mobile sellers' union to issue a trust badge to vetted shops. And I ran a new-and-used trade-in program for organisations' employees, with installment purchase. The site answered the buyer's question; those deals made sure there was something trustworthy on the other side of the answer.
What I killed
The idea I killed was buying growth. There was steady pressure to switch on paid acquisition and simply buy the traffic, and the numbers would have looked good in a review. I argued the other way: a marketplace that has to rent its audience never learns why people come. We stayed fully organic, which was slower and far less impressive to report, and it forced every gain to come from actually answering the user's question. I would make that trade again.
Outcome
Reached roughly 1.5 million monthly users, entirely through organic search, with no paid acquisition spend for the life of the product. The number I am proudest of is the zero next to the ad budget, because it means the growth was the product working, not a budget covering for it.
Where my role stopped
I owned product, design, and the partnerships. The engineering that made a million-plus pages fast and crawlable was not my code. I decided what those pages needed to say and prove, not how they were served. I also left before the marketplace side was rebuilt, so anything after my time is not mine to claim.
What I know now
I would have built the model pages a year sooner. I know exactly why I didn't: they weren't a 'feature', there was nothing to demo and no launch moment, so they kept losing to work that was easier to talk about in a review. The most valuable thing I built there was the hardest to put on a roadmap slide, and I have never forgotten how close it came to not getting built at all.