AEO tools and trackers
An AEO tracker checking once a month will show you a different set each month. 4 of the 5 question pools we measured were still growing at the fourth reading.
We repeated identical questions and counted what accumulated. Only 1 of 5 pools stopped producing new companies within 4 readings. That is the behaviour any tracker is sampling, and it decides what its numbers can honestly mean.
4 of 5
question pools still producing new companies at the fourth reading - the set had not settled
What we measured
- Only 1 of the 5 question pools we measured stopped producing new companies within four readings; the other 4 were still growing.
- On the broadest question we tested, the number of distinct companies named rose from 5 to 10 to 18 across 3 readings without repeating much of anything.
- A tracker sampling once a month is drawing one card from a pool of between 8 and 53 companies and reporting it as a position.
- Across 10 same-tier repeats of the same questions, two readings overlapped by 15% on average.
What settling would look like, and what we saw instead
A tracked metric is only a metric if the thing underneath it holds still long enough to be tracked. So we tested whether the named set settles.
It mostly did not. Repeating 5 identical questions, only 1 - the narrowest, “build me a new site” - stopped producing new names by the fourth reading, at 9 distinct companies. The other 4 kept finding companies they had not named before.
The clearest case is the broadest question: 3 readings, 5 companies each time, and 18 distinct names. Almost no repetition at all.
The consequence for a monthly chart
If the underlying set has not settled, a month-over-month line is drawing the sampling noise, not a trend. Two readings of one identical question overlapped by 15% on average in our repeats; the sharpest pair, taken 10 minutes apart on the same model tier, shared nothing.
None of this makes tracking pointless. It makes the unit of tracking a rate over many readings within a period, rather than one check per period. The difference is not cosmetic: it is the difference between a number that survives being re-run and one that does not.
It also means the honest chart has error bars, and most of the category's charts do not.
Who is cited when someone searches for an AEO tool
On the exact questions this page answers, the most cited domains are a marketing software company's blog, a software review directory and a forum, alongside a handful of specialist trackers. The review directory and the forum are the durable positions; the specialist tools rotate.
The questions this page answers
aeo tools
720/moaeo tracker
590/moanswer engine optimization tool
320/moanswer engine optimization tools
320/moaeo checker
140/mo
5 questions, 2,090 searches a month between them. All 6 answers we examined on them were produced by a web search, which means a page can reach every one.
Who an assistant cites on these questions today
- hubspot.com 5
- reddit.com 5
- g2.com 4
- tryprofound.com 4
- youtube.com 3
- aiclicks.io 2
- aeotrack.io 2
- airanklab.com 2
- aioseo.com 2
- scrunch.com 2
22 distinct domains are cited across these questions; the 10 above are the most frequent, with the number of answers each appears in. Counted from the answers captured in our corpus, not from a market census.
What to ask before you buy
- Ask how many readings sit behind a monthly figure. If the answer is one, the chart is sampling noise.
- Ask whether the tracker shows the raw answers. Every claim we make about this surface can be checked against the answer text, and so should theirs.
- Ask whether readings are separated by model tier and account type. Mixing them lets a change in the instrument look like a change in your visibility.
Honest limits of what we measured
- Five questions, one trade, one assistant, three to four readings each. Enough to show the set had not settled; not enough to say where it would.
- Capture-recapture assumes equal catchability, which is false here, so the pools are lower bounds.
- We have not tested any specific tracker. This is about the surface every tracker samples.
Questions we get asked
Does this apply to all AEO trackers?
It applies to the surface they sample. Any tracker taking one reading per period inherits the variance we measured; one taking many readings and reporting a rate does not.
How often should readings be taken?
More often than the reporting period, and we cannot yet tell you the right number - four of our five pools had not settled by the fourth reading. Anyone quoting a precise cadence has not measured this.
Is AEO tracking the same as GEO tracking?
The measurement is the same. The competitive set differs by which word the buyer types, which is why we keep separate pages.
What would you show us instead of a chart?
A rate with the readings behind it, the raw answers attached, and the model tier recorded per reading. Where the sample is too small to support a trend, we say so rather than drawing one.
Start with one free reading
One business, real questions, the raw answers back - before you decide on anything paid. If you are running this for a book of clients, say so and we will talk terms rather than quote a page price.
Free reading
Takes about a minute.
Related
- AI visibility tools
- LLM and AI search visibility tools
- AEO vs SEO
- Answer engine optimization
- All six measurements →
Pricing is per category, because the work differs by how competitive the answers are. See every category we measure and what a reading costs →
Published 31 August 2026. Every figure on this page is recomputed from raw observation files by a single script, so any of them can be traced back to the readings behind it. We make your site citable. We do not promise placement in any assistant’s answer - we measure whether you are in it, and show you the raw answers either way.