Here's a number that sent me down a rabbit hole. For cold outreach in B2B software, the 2025 industry average reply rate is somewhere between one and three percent. Top performers hit fifteen to twenty-five, but that's with tight targeting and long follow-up sequences. For most first-time indie founders, the realistic baseline is about one in fifty people writes back. Which means successful, on-benchmark distribution looks, to a developer's pattern-matching brain, indistinguishable from total failure.
The Teaching Signal Problem
In 2025 reporting, 48% of sales reps never send a follow-up email. Follow-ups generate 42% of all replies. Half the industry is quietly leaving nearly half their results on the table because the first email looked like it hadn't worked.
I don't think that's a sales problem. I think it's the same problem I'm having, and I think it has a specific shape.
I've been writing code professionally for close to a decade. Every meaningful skill I've built has been reinforced by a tight feedback loop. A compiler error is a signal in milliseconds. A failing test is a signal in seconds. A code review comes back in hours or days. The tool I use to update my internal model of whether I'm doing the work right is prediction error, and the dopamine circuit that handles it is a teaching signal, not a reward signal. What it teaches from is surprise. Short feedback loops are nutrient-dense. I've spent ten years eating them.
Distribution Switches That Signal Off
Distribution switches that signal off. You post on Indie Hackers and it sits for three hours with four views. You send twenty cold emails and get one reply, which is on benchmark, which feels like failure. You publish a blog post and, for a new domain, see essentially no organic traffic for three to six months because Google's sandbox is holding you in timeout. The feedback latency is so long, and the signal-to-noise so low, that the circuit trained on green checkmarks starts telling you, correctly from its own vantage, that nothing is happening.
So you course-correct. You rewrite the email. You try a new subreddit. You rethink your pricing. From the outside it looks like adaptive behavior. It isn't. You're course-correcting on a sample size of one and a measurement window three orders of magnitude too short.
The 12-Email Problem
I'm watching myself make every one of these mistakes in real time. Last month I wrote a cold outreach template, sent twelve emails, got zero replies in forty-eight hours, and immediately started rewriting the template. The honest read: twelve emails is not a sample. Zero replies in forty-eight hours is consistent with a 3% reply rate. The template might be fine. I don't actually know, because I didn't give it enough data to know. I rewrote it anyway.
Three Rules That Are Starting to Help
The frame that's starting to help, partially:
- Commit to a channel by volume, not by calendar. 100 cold emails before touching the template. 20 IH posts before changing the voice. Three months of weekly blog posts before grading the SEO bet. The threshold has to be high enough that noise averages out.
- Treat non-response as the baseline, not the verdict. A 97% non-response rate is what the channel looks like working. The 3% is your signal. Stop subtracting from a number that was never yours.
- Follow up. If 42% of replies come from follow-ups and half the industry doesn't send them, the single highest-leverage change most of us can make is the one we'd dismiss as a growth hack.
A Question Worth Asking the Community
None of this makes distribution feel good. I think the feel-good part only arrives once the data window is long enough for actual signal to come through, which for most channels is a quarter or more. The loop is long, the teaching signal is weak, the brain resists. That's not a character failure. It's a specification mismatch between how developers are trained and how distribution actually works.
If you've survived your first real distribution stretch as a solo dev, I'd like to calibrate on one number: what was the volume at which you finally stopped second-guessing a channel? Emails sent, posts written, DMs, threads replied to. Anything. I suspect most of us are quitting at n=5 when the right answer is n=100, and I'd love to hear where the truth actually falls. And if you're in your first 90 days right now and mostly hearing silence, that's also worth knowing about.
I write about building Flowly, a time tracker and invoicing tool for solo freelancers, as a solo developer learning distribution in public. This post was originally published on the Flowly blog.
Top comments (0)