| raw | raw/i-make-1m-year-hosting-pdfs-on-the-internet.txt |
|---|---|
| url |
TL;DR
Elston built Tiny Host (tiny.host) — drag-and-drop web/PDF/HTML hosting for non-technical people — to $1M ARR as a bootstrapped solo-to-small-team founder, while working a full-time bank job for the first 2.5 years. No reinvention: he modernized a 30-year-old validated category (web hosting), then won with boring, compounding marketing — SEO and faceless YouTube tutorials — plus relentless user-driven development. The vibe-coding wave (people generating HTML with Claude and needing to host it) is now blowing the product up.
The Founder’s Story
- Software engineer; quit a big US/London bank (“the only difference is which seat you’re in for the next 20 years”). Tiny was his 3rd–4th idea.
- Bought
tiny.host(two i’s) for $30 because the one-i domain was $300 — “shows how lightly I was thinking about this.” Built the prototype in 3–4 weeks deliberately as a marketing exercise (“I can build anything; I don’t know how to get users”).
Mistakes He Made First (and still sees everywhere)
- Spending 6 months building a B2C social app, launching to “crickets.”
- Too many half-committed co-founders; chasing YC/investors; holding the idea too tightly instead of building for the market; analysis paralysis.
- On the multi-product (Marc Lou / Peter Levels) approach: valid as a lifestyle, but “if you really want to go far, focus and go all in” — Dropbox wasn’t a side bet.
The Boring-Marketing Engine
- First users from Reddit (honest “I built this, what do you think?”), then Product Hunt for first paying customers. Monetized requested features by putting them behind a paywall (custom domains, password protect, QR codes).
- SEO is the foundation (~100K+ visitors/month). Process: Ahrefs/Semrush keyword research → target low keyword difficulty + real search volume → build landing pages (bottom-of-funnel action queries like “upload PDF”) + blog posts (informational) + YouTube around each use case. Plus domain-rating/backlink trust and technical SEO (Lighthouse, FCP, page speed). Now a 3-person SEO team (links, research, content).
- YouTube is underrated — faceless how-tos (“how to upload a PDF online”). Evergreen: some videos took a year to rank, then hit 100K views; highly converting (a how-to viewer is about to act); also surfaces in SEO.
- Jobs-to-be-done framing: nobody searches “web hosting”; they search “how do I upload my CV / restaurant menu / HTML file.” One universal product, many audience angles.
The Operating Philosophy
- User-driven development: observe behavior, ship lightweight (password protection shipped with no backend at first — manually set passwords until the real feature existed), make 100 tiny improvements that compound into a product that “feels like exactly what I needed.”
- Say no relentlessly to stay simple — deliberately not going deep like Vercel/Netlify; positioning as the DIY stack for AI-empowered non-technical builders.
- Founder–market fit > copying Steve Jobs: “aim to be successful, not Steve Jobs.” Find a mentor just ahead of you (you’re at $10K, they’re at $30–40K) via community — don’t ask “be my mentor,” just ask good questions and show progress.
- Bootstrapping = infinite runway. Like Tally and Postiz, a good product can sit and wait for a trend (here: vibe coding / HTML hosting) then skyrocket. Last year grew 3–4%/mo; this year 20–30%/mo.
Key Stats
- ARR: $1M · SEO traffic: ~100K+ visits/mo (peaked ~70K users/mo via SEO) · Signups: 2M+
- Left job at ~$8K MRR after 2.5 years; ~doubled yearly since · Team <5
Related Frameworks & Playbooks
- seo-growth — keyword difficulty, landing-page vs blog structure, YouTube SEO, technical SEO
- vibe-coding — the HTML-hosting tailwind from AI builders
- micro-saas — single universal tool, many audiences, say-no discipline
- idea-validation — “if you can find one customer, you can find 10, then 100”