How I Got 2,000 Users in 3 Months With Zero Ads

Everyone talks about "just ship it" but nobody tells you what happens after you ship and get 12 visitors on day one (8 of which are you refreshing from different devices).

I launched VectoSolve in late 2025. It's an AI tool that converts images to clean vector SVGs. Three months later: 2,150+ users, 50,000+ conversions, 140+ countries, and growing revenue with zero ad spend.

Here's what I learned (the useful parts, not the "believe in yourself" fluff).

The stack

Next.js 14 (App Router), Supabase (auth + DB + storage), Stripe (payments), Vercel (hosting). Total infrastructure cost under $50/month even with thousands of conversions. Nothing exotic, just solid boring tech that works.

I use Claude Code for 100% of the development. Not Copilot, not ChatGPT in a browser. Claude Code in the terminal. It writes, tests, and I deploy. That setup lets me ship features in an afternoon that would take a small team a week. If you're a solo dev and you're not using this, you're handicapping yourself.

Lesson 1: Price higher than your gut tells you

I started at $0.65 per conversion. Seemed fair, low barrier, easy impulse buy. Reality: Stripe takes $0.30 + 2.9% on every transaction. On a $0.65 sale, I was keeping $0.33. That's not a business, that's a rounding error.

Moved to $2.50 as the minimum entry point. Revenue per transaction went up 6x. Number of transactions dropped maybe 20%. Net revenue doubled overnight.

The people willing to pay $0.65 are almost always willing to pay $2.50. The only thing stopping you from charging more is your own impostor syndrome.

Lesson 2: Google Ads are a trap at low budget

Tried Google Ads at 3 euros/day for "png to svg converter." The clicks came but nobody converted. Why? People searching "free png to svg" want FREE. They click the ad, see a price, and bounce. I was literally paying Google to send me people who would never pay.

Killed it after a week. Organic SEO with good content has been 100x more effective and it costs nothing except time.

Lesson 3: Build for specific people, not "everyone"

The generic "image to SVG converter" market is dominated by tools with 10+ years of history and millions of monthly visitors. I couldn't compete head-on.

What I could do: build for specific communities that the big players ignore. Crafters who use cutting machines. Sellers on digital marketplaces. People who work with laser cutters. Each group has specific needs (specific file formats, specific optimization, specific vocabulary) that a generic tool doesn't address.

Speaking their language and solving their exact problem converts way better than being the 15th "upload image get vector" tool on Google.

Lesson 4: Over-engineering kills momentum

I built an SVG animation tool, a pattern generator, a brand identity kit. Cool tech. Zero usage. The features people actually pay for are dead simple: vectorize my image, remove the background, upscale the resolution. That's 90% of revenue.

Every hour spent on a fancy feature nobody asked for is an hour not spent on distribution, content, or talking to users. I learned this the expensive way.

Lesson 5: Distribution is the product

The best product with zero distribution loses to a mid product with great distribution. Every single time.

I spend more time on distribution now than on coding features. Publishing packages to ecosystems where developers hang out. Writing articles. Building integrations with tools my users already use. Making sure Google understands what my pages are about.

Code is easy (especially with Claude Code). Getting people to find your code and pay for it is the actual hard part.

The uncomfortable truth about solo SaaS

Nobody talks about the days where you make $3 in revenue and wonder if you're delusional. The checkout sessions that expire at 2am (someone almost paid but didn't). The support emails from confused users at 6am. The SEO changes that take 3 weeks to show any effect.

It's not glamorous. Most days it's just you, your laptop, and the quiet hope that the graph goes up a little more this week than last week.

But when someone from a country you've never visited pays for your tool at 4am while you're sleeping, something clicks. You built something real. It works. People find it, use it, and voluntarily give you money for it.

That feeling is worth every $3 day.

Try it

VectoSolve: https://vectosolve.com For devs (API, SDK, CLI): https://vectosolve.com/developers GitHub: https://github.com/Vectosolve

If you're building something solo too, I'm always down to chat. The indie dev journey is better when you share the war stories.