How we review other tools.
AEOCheck sells a product. Every comparison and review on this site is written by a company with something to sell, and the only honest way to do that is to publish the rules and let you check our work against them.
Last updated 2026-09-04 · Written by Tyson McKay, Founder, AEOCheck
The conflict of interest, stated first
We compete with most of the tools we write about. So every comparison and review carries a disclosure line, names at least one thing the other tool does better than we do, and includes a “choose them if” list with real reasons on it. If a page of ours reads like the other product has no case, that page is broken — tell us.
What we read, and what we will not infer
Every claim about another product is read from that company’s own published material: their pricing page, their help centre, their documentation, their announcements. We do not repeat what a third-party listicle says about them. The date we last read it is printed on the page.
Each row of a comparison table carries one of four marks, and the rule for the hardest one is strict:
- ✓ They do this, per their own material.
- – Partly, or it depends on the plan.
- ? Their material does not say. Silence is never treated as absence.
- × They do not do this — and we only print this when their own material says so, or when we have used the product and looked. A negative claim always states its basis.
Prices
Pricing in this category changes monthly. Every price we print carries the date we read it, and we re-check every page at least quarterly. A price we cannot read from the vendor’s own page is printed as “not published” or “price on request” — we do not fill the gap from a review site, and we do not carry forward a number from an older version of the page.
Nobody pays to appear
No tool has paid to be included, ranked, or described a particular way, and no page on this site carries an affiliate link. Links to other vendors are plain links marked nofollow, so that recommending someone costs us nothing and earns us nothing.
Our own numbers
When a page cites a statistic from AEOCheck’s scan corpus, the number is read from the database when the page loads, not typed in once. It is only printed at all when the sample is large enough to mean something (currently 30 businesses or more). And it comes with a caveat we would rather state than have pointed out: the businesses in our corpus are largely people who ran a check because they already suspected a problem. That is a self-selected sample, and it says something about people who came to us, not about every local business in the country. Where we make a claim about the market rather than our customers, it will come from a sampled frame we describe, or it will not be made.
Statistics we cite, and ones we refuse to
Every statistic on this site links to the primary source it came from — the study, not a blog that quotes the study — with the sample size and the date. Some numbers circulate widely in this niche without a traceable source or with a source that says something different, and we will not repeat them however often we see them. Among them: that Foursquare supplies most of ChatGPT’s local business data (rebutted); that businesses under about 150 reviews rarely get named by AI (no primary study); that search volume would fall 25% by 2026 (a 2024 prediction, not a measurement, and it has not happened). If you find one of these on a page of ours, that is a bug.
What the marks on our own product mean
On a comparison page, the AEOCheck column describes two things honestly separated: the $25 report, which is a one-time measured diagnosis with a prioritised fix list, and the done-for-you engagement, which is separately scoped and priced work. A row that only the engagement delivers says so. The report’s scores are a snapshot of probabilistic model output on a stated date, not a ranking and not a promise; the report says that on its first page, and so do we here.
Corrections
If we have a price, a feature, or a fact wrong, send it with a link to the vendor’s own page and we will fix it and update the verification date. Vendors are welcome to do this too — a correction from the company itself is the best source we can get.