Tag: #call-copilot

Nine months of Call Copilot in one breakdown: the full funnel from install to payment, €517 of ads that brought zero sales, and why the most active customer loses me money.

Overnight I lost a user who had made it through the whole funnel: Deepgram wouldn't let him record, and the extension showed a raw technical error. Working through the case, I realized the analyst agent that explained it to me could be a product in its own right.

A lead from the US got further down the funnel than anyone before him — and stumbled over microphone access. Two and a half hours on one email, written by hand — and the first hint of a profile of my target audience.

Shipped 1.9.10 with a fully simulated demo and fixed onboarding — and the stats flipped negative: almost 30 people gone in 3 days. Instead of chasing growth, I cut about 700 euros a year in costs.

Hooked up analytics and watched what users do with the extension. They press the record button five times, close the sidebar, uninstall. I'm redoing the onboarding and the demo.

Shipped a version with automated emails and a new listing, then dug in over the weekend and found it: since March, Stripe had been sending the renewal webhook differently, and people's credits weren't being credited. Sat up until three in the morning fixing it.

An AI review showed that almost nothing in the extension actually works. I fixed it, shipped two versions with mandatory registration — and deliberately stopped, to let the numbers accumulate.

I sat down to look at a dormant product: Supabase at 35 dollars a month, 75% of tokens going to free traffic, zero payback — and a hundred new users a month from organic growth. Deciding what to do with it.

I finally went through Call Copilot's install statistics and found out that three quarters of the tokens are burned by free and anonymous users. I switch off the free plan, introduce registration and a hard paywall — and can't properly test the thing I just built.

In one month transcription cost about 180 dollars, and free users burned three quarters of that. Meanwhile the fourth purchase in the project came in at 2 euros. I decide to shut the free plan down.

Six weeks with almost no movement: updated the paywall, signed up for an accelerator, and then the extension lived on its own for two weeks and suddenly brought in 24 euros. The third purchase — 1.99.

Ads are off, traffic is organic only — and on March 3, 2026 the first payment from a stranger lands in Stripe. Exactly what I'd been waiting for since December. And exactly one payment.

I finally built myself proper analytics and ran the funnel math: 10% activation, almost everyone sees the paywall, zero payments. The arithmetic says turn off the ads — so I'm turning them off.

Launched ads across ten Western European countries, and people uninstall the extension saying it's not what they expected. I dig into the numbers, move from ads to SEO, and find a competitor with three thousand users.

Found out that the extension's identifier is the same across all installs, not a new one each time — I went looking for workarounds for nothing. Rebuilt attribution and went to my first tracking session.

Launched Google Ads, set up a CRM, wired up emails. A decent number of installs, a euro apiece. And then the drop-off: people install the extension and do nothing.

The extension officially hit the Chrome Store, and the first payments came from friends. And it turned out one of my servers had been mining someone's crypto for a week — courtesy of a fresh hole in React Server Components.

Google's approval came in overnight, and the excitement was real — I'd been working toward it for a long time. Then: onboarding, lead forms, a stubborn Stripe, and turning down a fifty-fifty partnership.

Feedback from the launch course: the product works, but I'm not doing the one thing that matters — no label, no shelf, no ads. Figuring out how to fix that.

I bought Mikhail Tabunov's course on launching pet projects. The main idea of the first modules is surprisingly simple: don't build anything big — take an unfinished project you already have and get it to revenue.