How we score SEO tools
15 scoring criteria
Every public listing is scored on the same 15 criteria. Each score is a whole number from 0 to 5, backed by a one-line summary and an evidence note that names the vendor page or document we read. Several criteria also carry subcriteria, which are scored separately and shown on the tool page.
- Research depthDiscovery
- How well the tool supports keyword, competitor, SERP, backlink, or market research before execution.
- Keyword research quality for SEO contentDiscovery
- How well the tool identifies, qualifies, clusters, and prioritizes keywords specifically for SEO content creation.
- Content executionExecution
- How well the tool helps produce, optimize, or operationalize SEO content.
- Context gatheringGrounding
- How well the tool gathers the business, competitor, SEO, internal, third-party, and custom API context needed for differentiated recommendations and content.
- AI-search / GEOVisibility
- How directly the tool helps teams improve visibility in AI answer engines and citation-oriented search journeys.
- GEO readinessVisibility
- How ready the tool is for generative engine optimization: answer-engine visibility, citation likelihood, AI-search prompt surfaces, and content structures that AI systems can use.
- Technical SEODiagnostics
- How strong the tool is for crawling, audits, indexability, metadata, status codes, and technical site health.
- Publishing workflowOperations
- How well the tool moves work into CMS, publishing, approvals, or automated production workflows.
- Measurement & trackingMeasurement
- How well the tool monitors rankings, content performance, visibility, or ongoing SEO progress.
- Team collaborationOperations
- How well the tool fits multi-person workflows, agencies, approvals, handoffs, and repeatable processes.
- Value accessibilityBuying
- How easy it is for a team to reach value relative to pricing, setup complexity, and required expertise.
- Free toolsBuying
- How useful the tool is before payment through free tools, free plans, trials, or public utilities.
- Estimated customer scaleMarket proof
- Estimated customer or user scale based on public claims, with source claims stored separately for auditability.
- Agent readiness levelAutomation
- How ready the product is to operate as or inside an AI agent workflow, based on tracked subcriteria for autonomy, controls, integrations, memory, and evaluation.
- Ease of use and autonomous levelAutomation
- How easy it is for a team to reach useful outcomes, and how much work the product can perform autonomously after setup.
The 0 to 5 scale
We score the specific product, not the vendor or the category. A page-speed checker does not get keyword research credit because its parent company sells a keyword tool. Free access, open-source distribution, and broad adoption count only where a criterion asks about them.
- 0
- The feature the criterion describes is absent.
- 1
- Weak, narrow, or indirect support that exists but would not be a reason to buy.
- 2
- Basic support. It works for simple cases and needs other tools for the rest.
- 3
- Solid support for the common case. A reasonable buyer could rely on it.
- 4
- Strong. Among the better options in the category for this job.
- 5
- One of the strongest tools available for this job.
Category weight profiles
A rank tracker cannot publish articles and should not lose points for it. The overall grade on a tool page is a weighted average of its fifteen scores, using the weight profile of its primary category. A weight of 0 means the criterion is shown but not counted. Compare pages show each tool against its own category, and say so when the two tools do different jobs.
| Criterion | Technical SEO | SEO research | AI SEO automation | Rank tracking | Content optimization | Digital PR |
|---|---|---|---|---|---|---|
| Research depth | 2 | 6 | 3 | 2 | 2 | 4 |
| Keyword research quality for SEO content | 0 | 5 | 3 | 2 | 4 | 1 |
| Content execution |
What counts as evidence
We use the vendor's pricing page, feature pages, documentation, changelog, and public case studies. Review-site aggregates and press releases do not move a score. Pricing is quoted as shown on the vendor page on the day we checked, with the source URL stored on the listing and shown under the pricing table.
Draft listings stay out of public lists until an editor checks the pricing, scores the product itself, and records evidence for each score.
Re-verification
We check published prices every quarter. A script flags changed prices for an editor to review. Vendors can send corrections to hello@seotools.directory.
Editorial independence
Vendors cannot buy a rank or a higher grade.