The Content Optimizer scores one page, URL or draft against a target query and breaks the result into eight explainable sub-scores: Search Optimization, Intent Alignment, Semantic Coverage, Topic Coverage, Entity Coverage, Internal Linking, Content Gap Coverage and AI Citation Readiness. Each number is derived from a real signal, not a keyword-density gauge, and each comes with the exact work that would raise it. It is an analyzer and an editing assistant, not an article generator.
How it works
- 1Give a keyword and either a URL on the site or a pasted draft. Semantyra reads the page, pulls the live top-of-SERP for the query, and gathers every project signal it already has: the semantic and topical map, entities, search intent, keyword data, competitors, content gaps, Search Console, internal links, cannibalisation, AI citations and Project Knowledge.
- 2It computes the eight sub-scores. A sub-score reports "not available" rather than zero when the data to judge it honestly is missing, and the overall score is the weighted mean of the ones it could compute, renormalised over them.
- 3Each sub-score lists the specific actions that would raise it, weakest area first. Where the Fix Assistant can prepare the work - a section to add, a missing entity, an internal link, a page to create - it hands you the copyable pack.
- 4On paid plans an optional model pass rewords those actions for your page and can draft one requested section. It never changes a score or raises a new issue.
What you get
- A score you can defend line by line, with the signal behind every sub-score.
- A ranked, page-specific action list instead of a generic checklist.
- Prepared, copyable fixes for the actions the Fix Assistant covers.
- A re-score after edits that shows whether the work actually moved the number.
See how it works: Content Optimizer
Frequently asked questions
- Is this a Surfer or Clearscope style content grader?
- It overlaps on the SERP-comparison part, but the score is not term density. Most of it comes from your own crawl, semantic map, entities, internal links and Search Console, so it tells you what to do on this specific page, not just which words to add.
- Does it write the article for me?
- No. It analyses, scores and prepares the work. The optional model pass only rewords the prepared actions and can draft one section you ask for; it cannot invent a finding or change a score.
- What happens when a signal is missing?
- That sub-score says "not available" and is left out of the overall score, which is renormalised over the sub-scores that could be computed. Nothing is scored as zero because the data was not there.