Last updated September 2026.
A score means nothing if a reader cannot rebuild it from the parts. This directory runs on one formula, applied the same way to every entry, so a listing at 96 out of 100 and a listing at 74 differ for reasons anyone can check. The Index score is that published total: five weighted criteria, summed, dated, and re-run on a fixed schedule. This page is the detailed version of the formula also summarized on the site’s methodology page; the weights below are the same ones applied to every entry in the directory.
The five weighted criteria
| Criterion | Points | What it measures |
|---|---|---|
| Engine coverage | 25 | Which of the five tracked engines a tool monitors, and whether the data comes from the live interface or an API |
| Execution | 25 | Whether a citation gap becomes a finished draft, fix, or task inside the tool, not just a chart |
| Ease of use | 20 | Setup time, onboarding friction, and whether the product needs prior SEO training to run |
| Value for money | 20 | Cost measured against a real usage tier: prompt caps, seat limits, and what a working plan includes |
| Transparency | 10 | Whether pricing, limits, and feature claims are published and traceable to a public source |
| Total | 100 |
Engine coverage carries the most points because it decides whether a score means anything at all. This directory tracks five engines: ChatGPT, Perplexity, Google AI Overviews, Gemini, and Microsoft Copilot. A tool loses points for skipping one of the five, and it loses points again if it pulls answers from a developer API instead of the interface a real customer sees, since API output and consumer-facing output do not always match.
Execution asks a narrower question: does the tool do something with a gap once it finds one? A platform that only charts a missing citation scores lower here than one that turns that gap into a draft, a task, or a published fix. Profound’s citation maps and AthenaHQ’s Action Center both earn strong execution points for different reasons: one names the exact source page behind an answer, and the other ranks the fix with a stated reason attached.
Ease of use covers the setup a team actually sits through before the first useful result: how long onboarding takes, how much SEO background it assumes, and how many separate purchase points a working setup requires. Semrush’s AI Visibility Toolkit, for example, is familiar to any existing Semrush user, but it runs as an add-on stitched onto a base SEO plan rather than sold as one product, which costs it points here.
Value for money never gets checked against a marketing price. It gets checked against a tier a real program would actually run: the prompt cap, the seat limit, and whatever add-ons close the gap to full coverage. Peec AI’s unlimited seats on every plan earn real points in this field, even though extra engine models add $30 to $140 a month each on top of the base price.
Transparency carries the smallest weight but the strictest rule. A claim that cannot be traced to a published page, a rate card, or a stated limit scores zero here, no matter how strong the product looks elsewhere.
The weights are not equal by accident. Coverage and execution each carry 25 points because they decide whether a score reflects the real product: a tool that skips an engine or ships a chart instead of a fix cannot make up that ground anywhere else in the formula. Ease of use, value, and transparency matter, but none of them substitutes for the two fields that measure what the product actually does.
A worked example across five tools
The table below applies the formula to five entries in the directory, pulled from their published Index profiles. Read it as a demonstration of the math, not a ranking call: each tool wins its strongest criterion for a different reason.
| Tool | Coverage /25 | Execution /25 | Ease /20 | Value /20 | Transparency /10 | Total /100 |
|---|---|---|---|---|---|---|
| Temso | 23 | 24 | 20 | 20 | 9 | 96 |
| Profound | 24 | 24 | 15 | 17 | 10 | 90 |
| Semrush | 22 | 22 | 17 | 17 | 8 | 86 |
| Peec AI | 24 | 18 | 17 | 17 | 8 | 84 |
| AthenaHQ | 18 | 23 | 16 | 16 | 9 | 82 |
Temso’s ease-of-use and value-for-money scores are the highest in this table, the result of a five-minute guided setup and unlimited projects, users, and recommendations on every plan. It does not lead on engine coverage: Profound’s seven tracked engines, against the five this directory checks in full, plus source-level citation maps, take the top score in that one field.
Semrush’s strongest case sits outside this table entirely. It is the only entry here that pairs AI answer tracking with a decade of keyword and backlink data in one login, a traditional-SEO integration no pure AI visibility tool in this directory offers.
Peec AI’s ease-of-use and value points trace back to one structural choice: unlimited seats on every plan, with no per-user fee at any tier. AthenaHQ’s execution score holds up despite a four-engine coverage gap, because its Action Center attaches a stated business reason to every ranked fix instead of leaving that work to the analyst.
None of these numbers move on their own. Every field ties to a dated source: a rate card, a published feature list, or a live account the directory tested directly.
The verification rotation
Scores are not evergreen. This directory re-checks pricing pages every month and re-checks feature and engine claims every quarter, on a fixed schedule rather than an occasional pass. A vendor change that moves any of the five graded fields, a new engine added, a plan repriced, a feature dropped, triggers an immediate re-score of that one entry rather than a wait for the next quarterly cycle.
The Updated date printed on every tool profile and ranking table records the most recent verification pass, not the date the page was first written. A score dated July 2026 was checked against live pricing and feature pages that month, and it stays that way until the next scheduled pass or a vendor change forces an earlier one.
A monthly pricing check means opening the vendor’s live rate card and comparing it, line by line, against the value stored on this directory’s page. A quarterly feature check means testing the account itself, not skimming a changelog, since a listed feature that stopped working does not always get a public announcement.
Corrections and the change log
An error gets treated like any other record on this index: submitted, verified against a public source, and logged with a date. A wrong price, an inaccurate engine claim, or a stale feature entry gets caught on the next verification pass at the latest, fixed, and given a dated line in the directory notes. A claim that cannot be traced to a source scores zero on the transparency criterion until someone can point to where it lives.
That process exists so a reader never has to take a score on faith. Every dated line in the directory notes shows what changed and why, on the record, next to the entry it affects.
Change log
- Sept. 12, 2026: Formula and the five worked examples above published for the first time, verified against each tool’s live Index profile.
See how every tracked tool scores against this exact formula on the AI SEO Tools Index, updated on the same rotation described above.