The index, and the person behind it.
geoaeo tests the tools that get brands cited in AI answers — independently, hands-on, and with a hard line between what's paid and what's ranked.
Why this exists
Every digital agency on earth seems to have discovered “AEO” in the same six months, and most of the tools that arrived with the gold rush grade their own homework. I found that annoying. So I built the thing I actually wanted — an index that tests these tools the way you'd use them, run by someone with no reason to flatter any of them.
geoaeo is that index. It scores the software that decides whether your brand gets cited in ChatGPT, Google AI Overviews, Perplexity, Gemini and the rest — and the score is the only thing here that isn't for sale.
How I test
I don't review tools from their marketing sites. I ran every platform in the index myself, on live brand accounts, over 30 days — the same 25 real buying prompts put through each one, then a fair share of their recommendations actioned to see what actually moved.
That last part is where most of them fall down. Plenty of tools generate a confident to-do list; far fewer change anything once you do the work — and you only find that out by doing the work. The full method, including where I'm still calibrating, lives on the methodology page.
How the money works
geoaeo is reader-supported, in two honest ways. Some links are affiliate links, and a few vendors pay for featured placement. Anything paid is labelled PARTNER, every time.
Neither one buys a better number. A vendor can pay to be visible; they cannot pay to be ranked higher — those positions come from the test and nothing else. The day that line blurs, the whole thing is worthless, so I keep it bright.
Who's behind it
I'm Marcus Taylor. I founded Venture Harbour, where I've spent years building and running web tools (some more successfully than others) — which is mostly how I ended up caring whether AI engines can find any of them.
geoaeo is independent of every tool it covers. No board seat, no equity, no advisory cheque from anyone in the rankings — just a strong preference for knowing which of these things work before recommending them to anyone.
One caveat
The scores you see today are preview numbers while the 2026 benchmark settles. The engines move fast, and a score that pretended to three decimal places of certainty in a market this young would be a lie. As the data firms up, so will the numbers — and I'll publish the per-tool logs behind them.
If something looks wrong, tell me. The index is only useful if it's current.