Our first 100 customers came from a single Google Doc

Bootstrapped, no marketing budget, no team. I had built Pondscale — a small log-aggregation tool for solo and small-team developers — over six months of nights and weekends. The product worked. Nobody knew it existed.

I didn't have the appetite or budget to do what every devtool blog post tells you to do: hire a content writer, run paid ads, sponsor podcasts, crank out SEO articles. The math on any of those, for a $24/month product, was atrocious.

What I had was opinions. Specific, well-documented, sometimes-controversial opinions about how logging should work in 2024. I had been writing them into private notes for years.

The problem was getting those opinions in front of the right developers without spending money I didn't have.

I'd been a platform engineer for nine years, mostly at companies whose log infrastructure I disliked. When I left to bootstrap I built the simplest possible version of the tool I'd always wanted: ingest logs, search them fast, charge a flat fee, refuse to be enterprise-ready.

I launched on Hacker News with a lukewarm reception. 600 signups, 12 paid conversions, then nothing for three weeks. I'd budgeted six months of personal runway. After three weeks of silence I had five months left and no clear path to growth.

My partner suggested — somewhat sarcastically — that I "just write the stuff you yell at me about during dinner." I laughed. Then I sat down and wrote a 4,200-word public Google Doc titled "Why most logging products are wrong about logging." It had 84 footnotes citing real engineering blog posts, papers, and benchmarks I disagreed with.

I tweeted the link. I went to bed.

The first morning was bad. The doc had 38 reads. Six of those were probably me refreshing the tab.

But then someone from a fairly well-known engineering blog quote-tweeted footnote 31 and disagreed with it publicly. I'd anticipated the disagreement and had a longer response ready. I dropped that response into the doc as a new section, in real time, while their thread was still active.

This turned out to be the unlock. The doc was alive. It was getting better while people were reading it. It also had screenshots from inside Pondscale demonstrating, with real numbers, exactly the things I was claiming.

Within three days the doc had 14,000 reads. Two podcasts emailed asking me on. The Pondscale signup page — which was just the doc plus a Stripe link at the bottom — converted at 9.4%. I closed the laptop on day five with 78 new paying customers.

The near-failure had been the first morning. If I'd interpreted 38 reads as failure and stopped, I'd have missed everything that followed.

I kept the doc alive for the next eighteen months. New footnotes. New rebuttals to engineering posts I disagreed with. New screenshots from inside the product as I shipped features. It became a strange kind of public engineering journal that doubled as the most effective sales asset I had.

By month nine I had 480 paying customers and $11,500 in MRR. By month eighteen I'd crossed $100k MRR — $1.2M ARR — with no paid marketing of any kind. The doc remained the #1 source of attributable signups, surpassing all subsequent SEO content I tried to write.

The Pondscale homepage today is still effectively a polished version of that original doc. Every prospect who reads it for fifteen minutes converts at a rate I have never been able to replicate with any traditional landing page.

1. 1
   Opinions distribute. Features don't.

   A list of features is a commodity. A specific, well-defended opinion about how something should work is a magnet.

2. 2
   Write the thing you yell at dinner about.

   It's the highest-conviction content you can produce, and conviction is what cuts through the noise.

3. 3
   Make the asset live, not static.

   A document that updates in response to public conversation is a dramatically different artifact from a blog post that ages immediately.

4. 4
   Day one is not signal.

   38 reads on the first morning would have looked like failure to most analytics dashboards. The compounding started on day two.

5. 5
   For small teams, distribution is voice.

   You don't need a content team. You need one founder willing to say something specific and defend it.