We Reached 1,000+ Users: Ziva AI Tool Spread Through Word of Mouth!

In hindsight, it seems obvious: when a tool really works, people will talk about it. We built a specialized tool for a game making AI called Ziva to take on the full spectrum of problems that game developers encounter.

Ziva just crossed 1,000+ users, and almost all of that growth came without ads, without big launches, and without any real marketing push. It spread mostly through word of mouth.

What Ziva is (and why it’s different):

Ziva is an AI tool built for game development directly inside the Godot Engine. Within the editor, with full access to the game project, Ziva was built to give the AI the full project's context.

No mix-ups, no hallucinations. We originally created Ziva to make our own games at home, and anything that could hinder us for even a moment would have been unacceptable.

What it Actually Does:

Right now, Ziva goes beyond basic code assistance and operates at the project level:

• Understands your full scene tree and node hierarchy, not just individual scripts

• Maps relationships between nodes, signals, and scripts across your project

• Debugs issues using actual project context instead of isolated guesses

• Modifies or generates code that aligns with your existing architecture

• Navigates your project structure without needing repeated explanations

• Adapts to how your game is built rather than forcing generic patterns

• Helps trace issues across multiple files and systems, not just single scripts

• Reduces context switching by working directly inside your workflow

The difference isn’t just “AI that writes code.” It’s AI that can reason about how your game is put together.

Why it Spread without Ads:

We didn’t benefit from traditional ads.

The only thing we needed ads for was making sure our website actually showed up in search. Being a new domain called “Ziva” competing with a lot of unrelated results, we needed to make sure people could find us if they were looking.

Outside of that, growth came from two things:

1. We focused heavily on creating content around what we were building.

2. And we stayed extremely active with users.

We interact with people joining our Discord constantly, answering questions, fixing issues, and having real conversations in public. Sort of a combination between support and ongoing collaboration. We often make new features within days of user suggestions or grind all-nighters when edge cases are found.

Instead of pushing the product outward, we stayed close to the people already using it. If it worked for them, they shared it. If it didn’t, we heard about it immediately and fixed it.

Where Ziva is Now:

Ziva is still early since its launch.

But it didn’t start yesterday. We spent about a year refining its core capabilities before ever putting it in front of users.

Even now, the pace hasn’t slowed down. The founders (us) are still spending as much time as possible building, adding features, and responding directly to bugs that come from real users.

Final Thoughts:

The AI space is crowded. It didn't work for us to focus on attention, and paid acquisition strategies did nothing for us. Speculation can help boost a company's perceived value, but does nothing for its longevity.

If your tool can prove itself through what it actually does, early distribution can take care of itself. Of course, word of mouth only works if your product can survive first contact. Most will probably not.

Curious how others here think about this. Have you seen word of mouth work for your own products, or has it been harder to rely on than it sounds?