一年内将GEO工具增长至中五位数MRR

Indie hacking since 1995

I’ve always been a bit of an internet nerd and a builder.

My background is in business and growth, but I started my career very early. I first got online in 1995 and built my first website shortly after. Not long after that, I was already designing and developing websites for clients.

Over the years, I’ve built, scaled, and sold websites, mobile apps, and Chrome extensions.

In 2007, I turned one of my own projects into a real business. In 2013, I started one of the first ASO (App Store Optimization) agencies globally, which now employs a team of 15 people. In 2017, I launched an ASO SaaS product. It didn’t become what I originally expected, but I ended up selling it in 2021.

Then, in May 2025, I saw brands appear (or not) in AI-generated answers, and it was clear this would matter. And I wanted to prove to myself that I could build a SaaS and turn it into a truly large business. So, I decided to move fast. I found two people I really wanted to build with, and we started LLM Pulse.

I launched in July 2025 with two cofounders — Adrián Rojas and Esteve Castells. We help brands monitor and improve how they show up in AI-driven search platforms like ChatGPT, Gemini, or Perplexity.

It is fully bootstrapped. Our team is just three cofounders, which forces us to stay extremely focused on what drives value. We only work on things that have a direct impact on growth or product value. No roadmap for the sake of it.

Currently, we're at a mid-five-figure MRR, and it's growing steadily month over month, with growth accelerating over time.

Launching in two months

With a previous SaaS project, I spent too much time and money building the first version. I didn't want to repeat that here. So, we started building the MVP of LLM Pulse in May 2025 and launched it in July.

Our goal wasn't to build a perfect product, but to quickly validate that we could reliably measure how brands show up in AI-generated answers.

So we focused only on the core: building something that worked, gave us reliable data, and that people were willing to pay for.

Once that worked, we layered everything else on top.

As far as the stack, we kept the stack simple and pragmatic. And we avoided typical big cloud setups. We run most of our infrastructure on dedicated servers instead. It gives us more control and keeps costs predictable as we scale.

No playbook

The biggest challenge has been building in a space with no clear playbooks. Things change fast, and you’re constantly making decisions with incomplete information.

We had to be creative and make decisions from first principles, testing things quickly and iterating based on what worked.

We’ve also been very reactive. We don’t try to solve problems before they exist. We deal with them when they show up.

My advice for those in a similar position: Keep it simple, move fast, and continuously adjust as you learn.

Three advantages

Three things have been especially helpful.

First, staying very agile as a small team. We avoid unnecessary meetings and keep communication simple, which allows us to move fast and stay focused. But it's also a challenge. We had to intentionally define roles, priorities, and communication to maintain efficiency with just three people.

Second, doing a lot of demos. Demos have been one of our main sources of learning. Talking directly to users, seeing how they interact with the product, and building relationships has helped us improve much faster.

Third, experience. We come from SEO and have built products before. We’ve made mistakes in the past, some of them quite serious, and that helps us avoid repeating them.

That combination, speed, direct feedback, and experience, has been a big advantage for us.

A combination of growth strategies

We didn’t rely on a single channel. A combination of strategies working together drove our growth.

Our personal brands were a big early driver. We actively shared what we were building and our perspective on AI Search, attracting the right audience.

As I mentioned, we’ve also done a lot of demos. Demos have been key not only for learning but also for converting users and building early relationships.

Additionally, we’ve created free tools as lead magnets, organized webinars, worked on SEO (our background), and published studies to share market insights.

Timing has also helped. AI Search is growing very fast, and a real need exists to understand and measure its developments.

Of course, a product that works well and users genuinely like has made a big difference too.

Everything we do aligns with growth; every action pushes in the same direction.

Doubling down on SEO

Given our background, SEO was a natural channel for us, but we kept it very focused.

We started with a strong technical foundation: clean architecture, proper indexing, and an easily crawlable and understandable structure. We also use structured data to make our content clearer for search engines.

From there, it’s about relevance. We create content based on real questions we see in the market, not on volume, and we focus on building our brand and positioning within this space.

Less content, but more useful. That’s been enough to make SEO (and AI Search) work for us.

Fail and make mistakes

Here's my advice:

Forget about perfection.

Build fast, ship often, and don’t waste time worrying about problems that don’t exist yet.

Most things you’re afraid of won’t happen, and the ones that do, you’ll figure them out.

You will fail and make mistakes. That’s part of the game.

Just keep going.

What's next?

Going forward, the plan is simple: Build a great product, give great support, and have a lot of happy customers.

Beyond that, we want to:

Keep the team small

Move fast

Double down on what’s already working.

We’re aiming to build something big. That’s an important part of the motivation this time. And keep going. Keep building, keep improving, and see how far we can take it. And have fun.