So Who Is This Product Actually For

By Georg Malahov

Two weeks ago I finally launched ads targeting the EU — picked ten Western European countries, running on keywords like live transcribe and Google live transcribe. The first day went fine: CTR held up, the ads work, people come and install the extension. Then the strangeness starts. Some just install it and do nothing. And some uninstall — and in the "why" field they write the same thing: not what I expected.

That phrasing is what hooked me. The keywords, as far as I can tell, are decent, targeted enough. So it's not breaking at the entrance but somewhere at the seam between what a person imagined and what they actually opened. And I still don't know what exactly they imagined. I'm planning to simply write to the people who uninstalled and ask. Maybe that's where the key to product-market fit lies, because my own guesses are clearly wrong.

In parallel I sat down to go through the data — now that attribution is fixed, the numbers can be trusted. The picture turned out very sobering. The most active users are organic — the ones who found it on their own through the Google sidebar. The paid ones, especially from India, barely stir: they install and that's it. In other words, I was paying for traffic that's worse than the free kind. And a line of its own: over sixty percent of people hit the paywall and turn around.

The obvious conclusion is to change the source. I regularly get emails offering to promote the extension, price tag around two hundred euros — exactly what I spent on ads over the whole month. Except there it would buy SEO work, not burned-through clicks. On top of that I'm thinking about Product Hunt and an article. The daily ad budget is getting cut.

I also showed the extension to someone I know. Not much came of it — he's simply not the target audience, he has nothing to transcribe. And that's when I caught myself: I'm still showing the product to everyone around — friends, colleagues, whoever is at hand. But the only conversations worth anything are with the people who actually need it.

And at the end of the week I stumbled on a competitor. An extension almost identical to mine, launched in August 2025, three thousand users. Exactly the niche I'd considered mine — and quiet. The market turned out to be far more crowded than I'd pictured, and on top of that the cost of software in general is heading toward zero — anyone can build something similar today in a couple of weeks. That's probably the biggest failure of these weeks: I spent half a year building something someone else built earlier, faster, and has already put in the hands of three thousand people.

I'll admit: I'm starting to run a little low on energy. So the decision is this — let go. Let the product live its own life for a while, collect some reviews, stop endlessly tinkering with it. Talk only to the target audience, only to people, not to the interface. Meanwhile I'm switching to another project of mine.

Whether this bet is right will become clear in a couple of months: if without ads, on SEO and organic alone, the product climbs to those same fifty active users a week, it means all this time I was simply buying the wrong people, and I should have left paid traffic sooner. If it doesn't get there and the traffic collapses to zero, then the source isn't the issue — the issue is that I never figured out who this is for.

For now I'm off to write to the people who uninstalled — it seems to be the only source of truth I have left.