Q1 summary: $86 earned, 580 downloads, and one app killed

5 min read

Why I bother writing this?

Building apps has been a journey for me. Although the numbers are still somewhat embarrassing, I have no doubt that I'm making progress. A few months ago I would jump like crazy after getting the first customer (my journey in 2025: link).

Today I feel like the first customer is an easy part that I can repeat at any time - but scaling up to $100-1000 MRR is another challenge.

This process is both challenging and exciting for me, and I think it can be valuable for others too. Most of the content is written by people who have reached $10k, $100k or more, but in reality we just need to figure out how to achieve the next small milestone, one at a time.

I hope it helps anyone, or sparks some discussion, or least other builders at my level can cross-reference their metrics with mine.

Q1 numbers

  • Act Mate: $25 revenue (Feb), 349 product page views, 199 downloads → killed
  • Hanzi Flow: $97 revenue (Feb), -$36 (Mar refund) = $61 net
  • Total: ~$86 revenue across both apps

Killing Act Mate

Act Mate was a cool app idea, that arose from a real problem proven by the discussions with actors (who need to learn lines faster).

The app also had some kind of ASO potential - we got into top 15 in the US for the keyword 'script rehearser'. This gave us 349 page views and 199 downloads over 2 months. These are quite good numbers and I think that if we nailed onboarding and conversion, we could have hit some recurring revenue with zero marketing efforts.

Why we killed it? AI!!!! The app was heavily based on AI features, which made it non-deterministic. Users loved the idea but complained about the execution because the product couldn't deliver a consistent experience. I think it was technically feasible to improve it, but we thought the effort was not worth it.

Two takeaways:

  • for side projects it's better to build SIMPLE stuff
  • ASO is a great way to grow on autopilot but you should research keywords IN ADVANCE when validating an idea, not after building the app

Hanzi Flow: TikTok as the growth engine

At the beginning of January I launched a new app: Hanzi Flow (AI flashcards for Chinese language). Hanzi Flow app The goal was simple: build a freemium app, deliver real value to actual users. Some metrics:

  • 580 downloads,
  • 441 users,
  • 15 trials,
  • 4 paid (1 retained) -> $61 net revenue

This time I decided to focus on TikTok marketing (because it's the only low-cost and high-scale distribution channel). Here is how marketing went: 51 TikToks, 225K views, 350 followers, 13.5K likes. First-time downloads chart Now I think that short-form marketing is just a game of numbers. Every Tiktok must have strong hook (to get views) and strong CTA (to convert views into app downloads). The more you post, the more data you have. Consistency compounds and turns bets into evidence-based decisions. I wanna figure it out and this will become my focus for April-June.

Hanzi Flow: Turning point in conversion

At the beginning of February, after getting a sensible number of users, I decided to work on conversion. I almost did not change the app functionality. But I highlighted premium features in the app and redesigned onboarding. v1.7 conversion comparison The result:

  • Before v1.7: 302 downloads, 1 in-app purchase
  • After v1.7: 277 downloads, 18 in-app purchases (important: in-app purchases also includes free trials that didn't convert!)
  • Same traffic, 18x more conversions.

I think it's an interesting learning.

Hanzi Flow: retention reality check (PostHog analytics)

The retention is not my problem at this moment, because I am still working on top of the funnel: user acquisition (marketing) and activation (onboarding, conversion into trials). But since my original assumption was to build something that people actually use, here are the results:

  • 441 total users, ~40 weekly active users consistently using the app

I don't have any retention-focused features (notifications, motivations, streak reminders etc) and I'm very happy someone really finds my app useful.

What's next: marketing mode for Hanzi Flow

The app is not perfect, and I do feel strong FOMO that other people are building cooler stuff, but I realized that the highest-leverage thing I can do is to learn how to make 10-100k people download your app. Once I nail this, I can improve Hanzi Flow, or even start a new project.

I'm giving myself 3 months, during which I will focus only on marketing. I have already built a custom AI toolset: a set of SKILLS for Claude Code that analyze tiktok competition, manage assets and create videos and slideshows on autopilot. I'll start with 5 accounts.

To sum up: the product that converts is better than the product that's perfect. Mine already converts, so now I have to figure out how to make it recognizable.