How we judge

Most tool reviews are rewritten feature pages. Ours are research syntheses. We do not run hands-on lab tests, and we do not pretend to — what we do is read everything of substance about a tool, weigh it, and score the result against a rubric that is the same for every tool in its category. The judgment applied to that evidence comes from an operator who runs SEO systems in production daily. This page is the rubric, and the evidence standards behind it.

The evidence base

Every report draws on four kinds of sources:

Vendor documentation and pricing pages — read critically. Vendor claims are treated as claims, not facts, and reports say so when a statement rests on one.
Release notes and changelogs — shipping history says more about a tool's direction than its roadmap does.
The documented experience of practitioners — public reviews, community threads, and published case studies, weighed for independence and recency.
Independent tests published by others — credited when we rely on them.

How evidence is weighed

Corroboration across independent sources outweighs any single glowing or scathing account.
Recent evidence outweighs old. Tools change monthly; a two-year-old complaint may already be fixed.
Where credible sources conflict, the report surfaces the conflict rather than averaging it away.
Where the evidence is thin, we say so — or we don't publish a score at all.

The scoring rubric

Evidence scores run 0–10 and are our editorial judgment of the documented evidence — not a lab measurement. Each score is the weighted result of five dimensions:

Output quality — does what the tool produces (keywords, briefs, audits, content scores) hold up when operators act on it?
Data accuracy — how the tool's numbers hold up against independent checks reported by practitioners and third parties.
Workflow fit — how much manual glue a solo operator or small team needs around the tool.
Value — what the useful tier really costs once limits and add-ons are counted.
Trajectory — release cadence and whether the vendor is fixing the right things.

What we will never do

Claim hands-on testing we have not done. This site is research synthesis, and it says so wherever it matters. Fabricated first-person anecdotes are the disease this methodology exists to avoid.
Invent statistics, benchmark numbers, or user counts. If a number appears in a report, it is sourced — and if we cannot verify it, we do not publish it.
Let commissions set a score. Affiliate relationships are disclosed on every monetized page; scores come from this rubric.
Hide who we are. This site is operated by the maker of RankFlywheel, an SEO automation platform. Where RankFlywheel appears in a comparison, that ownership is stated in the report itself, and we tell you plainly which kinds of buyers it is wrong for.

Updates and corrections

Tools change monthly. Reports carry a visible updated date, and we revisit a report when a vendor ships a change that could move a dimension score. If we got something wrong, the correction is noted in the report, not silently edited.