[13pixels]

Interesting tool. We've been looking at this space (GEO) as well.

One big challenge we've hit is the lack of determinism. Unlike Google where rank #1 is fairly stable, ChatGPT/Gemini answers vary wildly based on user history, location, and random seed.

How do you define a 'rank' in a conversational interface? Is it just frequency of mention over N runs? Or do you try to map 'top' recommendations to positions?

[gab4hikoo]

Great question, this is honestly one of the core technical challenges we've had to solve.

You're right that determinism is fundamentally different from traditional search. Our approach is statistica and we measure frequency of mention + positioning within the response.

So it's more like a visibility score built from hundreds of data points across platforms. Think of it as share of voice in AI responses rather than a fixed SERP position.

We also track the variance itself, some brands get consistent mentions regardless of prompt phrasing, while others are highly volatile. That variance is actually a useful signal.

Would love to compare notes on what you've been exploring in the space. Feel free to reach out (email in profile).