Best sales day ever, three days after my "$27 a week" post
Last week I posted "My landing page converts at 6.5%. I still only make $27 a week." This is what happened since, because the whole point of that post was the numbers.
Three days after writing it I had the best day in the app's history: 8 sales in one day. Previous record was 3.
What did it: two things landed at once. A write-up on MacMagazine, the biggest Apple site in Brazil, and a post on HN. Three of those sales were from Brazil — so I spent the next morning adding BRL pricing and Pix, and translating the delivery email into Portuguese. If a country buys, meet it where it pays.
The thesis of the original post held up embarrassingly well. The blog written in a country's own language converted. TikTok still does not appear in my referrers. Not low — absent.
Two more things I fixed since, both found by looking instead of guessing:
My homepage was serving 46 words to crawlers (React SPA). My refunds page outranked my actual homepage on Google for my own brand name. Prerendered the content into the HTML: 46 words → 1,865. If your landing page is an SPA, curl it once. You might meet your refunds page.
And version 2.3 shipped this week, built almost entirely from two users' bug reports. One of them had told me "I clicked the notifications and nothing happened" — turned out macOS eats the first click on a non-frontmost app. The fix took a day. Finding out it was broken took a stranger being kind enough to write.
Still $27-ish weeks between the spikes. But the machine is starting to look like a machine.
The strongest part is how quickly the founder turned actual demand signals into product changes. The Brazil sales, Pix adoption, and user bug reports all became concrete evidence rather than assumptions.
Kind of you, but I'd push back on "quickly". Brazil was an accident. Sales appeared and I went looking for why afterwards, which is the opposite order.
And the first-click bug had been broken since the day I shipped. I had no way of seeing it, since I collect no analytics at all, so it stayed broken until a stranger was decent enough to write and tell me. That is eight months of every new user hitting it.
Fast is what it looks like from outside. From in here it is mostly finding out late and then moving.
That’s a useful correction. The “finding out late” part may actually be the more interesting pattern here.