Entity Salience Checker: Google NL API v1, Page vs Target Entity
You know what your page is about. The question is whether a machine reading it agrees. This free Colab notebook sends your page’s main text to the Google Cloud Natural Language API, ranks every entity it finds by salience, and checks where the entity you meant to write about actually lands, compared with competitor pages on the same topic.
It’s a free, lite version of one check from the Clarity step of the 5C framework in my course The AI-powered Entity Audit.
Who it’s for
Entity-focused SEOs, content strategists and anyone auditing what a page is about to a machine, especially when a brand name, a newsletter box or a handful of synonyms might be pulling attention away from the topic.
How to use it
- Run the sample first. Three fictional pages about the French press, with saved sample API responses, run without a key.
- Add your API key. Enable the Cloud Natural Language API, create a key and store it as a Colab secret. The notebook never prints or saves it.
- List your pages. Add your URL and competitor URLs, the target entity, any aliases you use, and optionally its Wikipedia URL.
- Check the cost estimate. The notebook counts billable units before calling the API and caches every response.
- Read the clarity check. For each page you see the top entity, the target’s rank and salience, and a lite 0-3 clarity score with tips.
- Compare. See your target salience against competitor pages, and your main article text against the full page with navigation and footer.
- Export. Download the Sheets-ready workbook.
What you get
- Top entities per page by salience, with type, mentions, and Knowledge Graph MIDs and Wikipedia URLs when the API returns them
- Where your target entity ranks, and whether its attention is split across synonyms the API treats as separate entities
- Boilerplate and brand entities that outrank your topic
- Plain-English tips: mention the target early, use it in the title and headings, and pick one unambiguous name
- A comparison with competitor pages, plus CSV and Excel exports
Limits and costs
The notebook calls API version v1 on purpose: salience is only returned by v1, and the v2 entity object has no salience field. Billing is per unit of up to 1,000 characters per request, with 5,000 free units a month for entity analysis, so a few pages cost little or nothing. Salience is the Natural Language API’s own measure. Google hasn’t said Search or its AI features use it, and LLM-based assistants don’t call this API, so use it as a diagnostic rather than a ranking score.
For why template elements win, read Beatrice Gamba’s Why Navigation Elements Outrank Your Brand in Entity Analysis. The full audit, across Clarity, Coverage, Consistency, Connection and Corroboration, is in The AI-powered Entity Audit.
What’s included
- Extracts each page's main text and runs Google Cloud Natural Language API v1 entity analysis
- Shows where your target entity ranks by salience, and if synonyms split it into separate entities
- Flags brand, newsletter and other template entities that outrank your topic
- Compares your page with competitor pages for the same target entity
- Gives a lite 0-3 Clarity score with tips, plus a Sheets-ready export
Where this resource is used on MLforSEO
Walkthroughs on the blog
Concepts behind it
More entity SEO resources
- No-Code TemplateTopical Map Template and Generator Based on the EAV Model (Google Sheets)
- ChecklistGenerative Engine Optimisation (GEO) Checklist
- ChecklistKnowledge Panel & Brand SERP Audit Checklist
- Coding Script or NotebookStructured Data Automation Kit: Audit, Generate and Validate JSON-LD with Python
- WorkbookEntity Confidence Recovery Tracker
- No-Code TemplateEntity Analysis with Google Cloud Natural Language API (No-Code Template)
See all 18 entity SEO resources →See all 17 content analysis resources →See all 42 coding script or notebook resources →

