My founder emails, week one: 86 sent, almost nothing back
Last week I promised to report what came back from the three emails AppX sends from me. 86 emails, 58 people, zero replies. Here are the numbers and what I think they mean.
Ilyos Olimov · · 5 min read
Last week I wrote about the three emails AppX sends from me, and admitted that the reply address had been broken the whole time. I fixed it on October 21 and promised a follow-up with what came back.
This is that follow-up. The short version: almost nothing came back. I think that is the useful part.
The numbers
From October 17, when the first email went out, to October 26:
| When it goes out | Sent | |
|---|---|---|
| "Your first app is ready" | Within 15 minutes of your first successful build | 30 |
| "Your app is still where you left it" | You built an app, then nothing happened for 3 days | 28 |
| "Your 100 credits are waiting" | A day after signup, if you never started an app | 28 |
| Total | 86 |
The 86 emails went to 58 people. 27 of them got more than one. The server logs since October 20 show no failed sends.
What people did after
| What happened next | |
|---|---|
| First app ready (30) | 2 came back on a later day |
| Still where you left it (28) | 0 came back within a week |
| Credits waiting (28) | 1 started an app, and the build worked |
Nobody paid after any of the three.
The first email looked much better at first. 8 of the 30 kept building after it went out. Then I checked the times: the email arrives within 15 minutes of the first build, so those 8 people were still on the page. They did not come back because of the email. They had never left. Counted as "came back on a later day", it is 2 of 30.
Replies
Since October 21, when replies started reaching my inbox: zero. I searched by subject and for my own sign-off, in case someone replied with a new subject. Nothing.
Did the emails do anything?
I don't know, and I want to be precise about why.
To know whether an email works, you compare people who got it with similar people who didn't. Before October 17 nobody got these emails, so I looked at similar people from September and early October:
- People who built an app and then went quiet for 3 days: 0 of 4 came back within a week.
- September signups who didn't start an app on day one: 1 of 8 started within a week.
Those groups are tiny. September was quiet, and from October 1 to 16 nobody could build without paying, so there are almost no comparable people. What I can say: the emails did not visibly change anything. Before the emails, people who left mostly stayed gone. After the emails, they still do.
A correction
Last week's post said the "still where you left it" email asks "what stopped you?". It does not. It names your app, says how many days ago you built it and how many credits you have left, suggests one small change, and links to the app. Only the signup email asks: "If something stopped you, reply to this email." It got no replies either.
What I think is going on
Earlier this month I looked at the first week of every project. People who asked for an app with 7 or more screens mostly never sent a second message: 18 of 25. With 1 to 4 screens, it was 4 of 16. The bigger builds didn't fail more often. They took longer, around 15 minutes against 11, with nothing on the phone preview until the very end. Those numbers come from our older build records, 64 projects, so treat them as a hint.
If that is right, people decide during the build, not three days later. By the time an email arrives, the moment is gone. An email can remind someone that their app exists. It cannot give back the 15 minutes they spent watching a progress card.
So the next experiments are in the product, not the inbox: telling people up front how long a build takes and that they can leave, and showing something on the phone while it builds. The emails stay on. They cost nothing, and if one of them ever lands at the right moment, that is worth it.
What I would tell another founder
- Measure "came back on a later day", not "did anything after the email". My best-looking number, 8 of 30, was people who never left.
- An email cannot fix a first session. If people leave during their first use, look at that first use.
- 86 emails is not an experiment. Without a comparison group of real size, "it didn't work" and "nothing would have worked" look the same. Say which one you actually know.
I will report again when there is something new to report, including if someone finally replies.
Keep going
More practical ways to move your app forward.
- How to describe your app to AI: a prompt template
Write a clear AI app-builder prompt with a copyable template, a practical example, acceptance checks, and focused follow-up edits.
- How to turn a client brief into a mobile app prototype
Turn a client brief into a mobile app prototype with a copyable AppX prompt, a practical test checklist, and a focused client-feedback workflow.
- The 5-backtick fence: stopping prompt injection via Markdown
We ship user input into an LLM architect prompt every turn. A reviewer caught that a 4-backtick string from the user could close our 3-backtick fence and run arbitrary instructions on the model. The fix is one weird Markdown trick.
- Investigators before builders: a 2-phase LLM workflow
Most teams dispatch coding agents straight to the build. We run a read-only investigator phase first — and the findings have rewritten our plan in 30-40% of cases. Here's the shape, the brief format, and what to do when findings contradict the plan.
- Step 0: delete dead code before you refactor with an LLM
Every byte you load into an LLM agent's context is a byte it has to keep track of. Dead code looks like signal — and agents preserve, extend, or mimic it. Our rule: a pure-deletion commit before any structural refactor of a file over 300 LOC.
Ready to try your idea?
Start with a clear description. AppX will help you turn it into an app you can test.