Your next SEO move
- Create one AI SEO buying page showing a real, human-reviewed research workflow.
- Investigate the backlink checker’s low click rate before changing its promise or building more tools.
- Protect the homepage and MCP page: live checks place them first and fourth for their respective queries.
Recommendations
1. Give AI SEO buyers a page that demonstrates the product
Do this
- Create /ai-seo-tool with one reproducible keyword-to-report example and its source data.
- Show human review, supported assistants, setup, and a realistic monthly workload at the advertised price.
- Link it from the homepage, MCP feature page, and the Claude SEO guide.
Why
- The proposed URL returns 404; the sitemap has no dedicated AI SEO buying page.
- The existing AI-visibility page covers brand mentions, a different job from running research with an assistant.
- Tool evaluators plausibly fit OpenSEO’s offer; this is inferred fit, not measured conversion evidence.
- New rankings could bring qualified visits, but roundups dominate and the achievable click share is unknown.
| Query | Estimated US searches/month · Sep 17 research | OpenSEO visibility · Sep 20 live check |
|---|---|---|
| ai seo tool | 2,400 | Not in the first 20 results |
| ai seo tools | 2,400 | Not in the first 20 results |
The two volumes overlap and are not added. Depth 20 returned 17 and 16 organic listings respectively; “not in the first 20 results” means absent from that returned result set, not proof of deindexing.
- Search Console also shows zero clicks for both terms; the homepage received only 9 and 15 impressions.
- Its average positions were 52.7 and 62.9 across countries, August 20–September 17, 2026.
Keep AI Visibility focused on mentions; use the MCP feature page and the Claude SEO guide as supporting paths. A product page is plausible: SEO.AI appears fourth for the singular query, while One Little Web’s comparison leads both variants.
2. Diagnose the backlink checker’s click gap, then test one change
Do this
- Split the head query by country, device, and week in Search Console.
- For stable segments, test a title and search description emphasizing “top 15” and “no signup”.
- Keep the free checker distinct from the paid backlink-analysis feature; do not merge either page.
Why
- The head query produces many impressions but few clicks; it deserves a bounded experiment.
- A live US check did not reproduce the aggregate position, so a simple snippet explanation is unproven.
- More relevant clicks could feed trials, but neither conversion rate nor attainable traffic gain is known.
| Page / query | GSC clicks / impressions · Aug 20–Sep 17, all countries | What this establishes |
|---|---|---|
| Free checker / backlink checker | 727 / 1,599,818 | 0.045% calculated click-through; average position 9.0 |
| Backlink feature / backlink checker | 30 / 4,800 | Two URLs appear; overlap alone does not establish harm |
What else we checked
| Opportunity | What we found | Decision |
|---|---|---|
| Semrush comparison | 1,300 estimated US searches/month; proposed page absent. | Runner-up: clear buying intent, but AI research better expresses the demonstrated product workflow; comparison needs verified tradeoffs. |
| MCP / open-source winners | MCP #4 (page 1); homepage #1 (page 1). | Preserve intent; link to the new buying page, not a broad rewrite. |
| New discovery tools | Finder and generator now return 200; finder absent in live sample. | Measure before rewriting; Sep 18 suggestions have already become real pages. |
| Blog, library, docs | Two samples per family explain distinct workflows and setup. | Use relevant internal links; no mass-content defect established. |
| Maintenance | 50 healthy responses; 16 informational flags; HTTP/www remain 200. | Consolidate URL variants and tidy titles/descriptions; no demonstrated loss puts these first. |
How this report was made
Generated by the OpenSEO SEO Audit skill, run by Codex on September 20, 2026.
OpenSEO supplied the 50-page crawl, Search Console rows and live US English results. Recent September 17–18 keyword and backlink research informed discovery.
HTTP reads checked page content, status, links and canonical tags, which declare each page’s preferred URL. Tool execution, rendering, performance and conversion outcomes were not measured.
Evidence and methodology
Search observations
Organic positions count unpaid listings only, excluding ads and answer blocks. Both singular AI-tool checks found OpenSEO absent. Other results changed slightly between checks; these are snapshots, not trends. Search Console averages cover all countries and are not equivalent to live US positions.
| Query · US/en, depth 20 | Organic listings returned | OpenSEO URL and organic position | Retrieval date |
|---|---|---|---|
| ai seo tool (Initial) | 16 | Not seen | Sep 21 UTC / Sep 20 Pacific |
| seo mcp (Initial) | 16 | 4: https://openseo.so/features/mcp | Sep 21 UTC / Sep 20 Pacific |
| open source seo tools (Initial) | 18 | 1: https://openseo.so/ | Sep 21 UTC / Sep 20 Pacific |
| semrush alternatives (Initial) | 16 | Not seen | Sep 21 UTC / Sep 20 Pacific |
| ai seo tool (Final) | 17 | Not seen | Sep 21 UTC / Sep 20 Pacific |
| ai seo tools (Final) | 16 | Not seen | Sep 21 UTC / Sep 20 Pacific |
| backlink checker (Final) | 19 | Not seen | Sep 21 UTC / Sep 20 Pacific |
| competitor keyword finder (Final) | 17 | Not seen | Sep 21 UTC / Sep 20 Pacific |
Coverage and technical checks
The sitemap lists 91 URLs: product/features, free tools, pricing/company pages, blog, strategy library and setup/skill docs. No dedicated vendor-alternative or competitor-pricing family was listed. Fifty crawled pages returned 200; the crawl found six long titles, seven long descriptions and three short descriptions, all informational. No critical or warning issues were reported in that sample.
Fourteen existing sampled pages returned 200 with self-referencing canonical tags and no robots meta restriction found; the two proposed URLs returned 404. Robots.txt allows crawling. HTTP and www homepage variants return 200 rather than redirecting; no search loss is established. Curl could read the pages after another HTTP client was blocked, so this is not diagnosed as a public outage.
Content families read: homepage, pricing and open-source positioning; MCP, AI visibility and backlink features; backlink checker, competitor keyword finder and keyword generator; open-source tools and Claude SEO blog posts; GSC discovery and AI-assistant library guides; MCP and agent-setup docs. The free tools expose challenge text in static HTML; their interactive results were not tested.
Evidence provenance and sizing
Search Console: sc-domain:openseo.so, web search, August 20–September 17, 2026, query/page dimensions; 70 leading rows plus 30 AI SEO-filtered rows, not a complete export. The backlink head-query click rate is 727 ÷ 1,599,818 × 100 = 0.0454%. No impressions-to-visits forecast is made.
A current 25-row organic domain sample reported 166 recorded keywords; its sampled ranking dates differ from retrieval dates and did not substitute for live checks. September 17 research supplied AI-tool volume 2,400 per variant and Semrush-alternatives volume 1,300; September 18 supplied finder 480 and generator 1,900. These are provider estimates, not measured visits, and close variants overlap. September 17 backlink research recorded 138 referring domains; that historical count alone does not prove a ranking cause.
Sources
- https://openseo.so/pricing
- https://openseo.so/features/mcp
- https://openseo.so/features/ai-brand-visibility
- https://openseo.so/features/backlink-checker
- https://openseo.so/backlink-checker
- https://openseo.so/competitor-keyword-finder
- https://openseo.so/keyword-generator
- https://openseo.so/blogs/best-open-source-seo-tools
- https://openseo.so/blogs/how-to-prompt-claude-code-for-seo
- https://openseo.so/library/keyword-research/gsc-programmatic-discovery
- https://openseo.so/library/ai-agent-seo/run-seo-from-your-ai-assistant
- https://openseo.so/docs/mcp
- https://openseo.so/docs/agent-setup
- https://openseo.so/open-source-seo
Sitemap · Robots directives · AI-tool result leader · Vendor comparison example