- The score has four visible parts: topic coverage 45, length 20, keyword placement 20 and subheading topics 15. Nothing is hidden.
- Every target is counted from the comparison pages' own text. A topic is recommended only when several of them use it.
- Comparison pages come from the URLs you add, the results MrKeyword has seen in search, or its own predicted results. Each is labelled.
- Your own website is never compared with itself, and fetched pages are cached for 30 minutes so re-scoring an edit fetches nothing again.
- It does not write the article. It shows what the ranking pages cover and what your draft is missing.
What the Content Optimizer does
The Content Optimizer answers one question: how does this draft compare with the pages that already rank for the keyword? It does not guess what Google wants. It reads the pages in the results and counts what they have in common.
You can paste a draft, with ## marking subheadings if you like, or give the URL of a live page. Then you name the keyword and, optionally, the pages to compare against. If you leave the comparison URLs empty, MrKeyword uses the results it has already observed for that keyword, or its own predicted results, and labels which one it used.
How the score is built
The score is out of 100 and is the sum of four parts, each shown with its points and a short note:
| Part | Points | What it measures |
|---|---|---|
| Topic coverage | 45 | The share of the topics the ranking pages share that your draft also covers. |
| Length | 20 | How close your draft is to the middle range of the comparison pages' word counts. |
| Keyword placement | 20 | Whether the keyword appears in the title, H1, opening and body, without being stuffed. |
| Subheading topics | 15 | How many of the words several ranking pages use in their H2 and H3 headings appear in your subheadings. |
Each part is calculated from the comparison pages, not from a fixed checklist. Change the comparison pages and the targets change with them.
Topics, uses and subheading ideas
Below the score, a table lists the words and phrases that enough comparison pages use, how many of them use each one, their average number of uses, and how often your draft uses it. A topic is marked missing, covered or overused. Overused means your draft uses it far more often per 1,000 words than the ranking pages do.
A second list shows subheading ideas: words that appear in the H2 and H3 headings of at least two comparison pages. They are prompts for sections a reader might expect, not a template to copy.
The comparison table names each page, where it came from, its word count and its number of headings. Pages that could not be read are listed with the reason, so a failed fetch never quietly changes the score.
Where the comparison pages come from
- You added the URLs yourself, one per line.
- Seen in search means MrKeyword has observed results for that keyword in its own search data.
- Predicted means the results come from MrKeyword's own index, and are labelled as predictions.
At least two readable comparison pages are needed. Your own website's pages are left out, so the draft is never scored against itself. Fetched pages are cached for 30 minutes, and a workspace can run up to 60 analyses an hour.
What the Content Optimizer does not do
- It does not write or rewrite the article. It shows what the ranking pages cover and what your draft is missing.
- It does not promise a ranking. It compares text, not links, authority or search features.
- It does not use search volume or a difficulty score. The targets come from the comparison pages' text.
- It does not read pages behind a login, a paywall or a block. Those pages are listed as unreadable.
- It does not compare your draft with your own site's pages.
Content Optimizer in each plan
The optimizer is available on every plan. It is limited to 60 analyses an hour per workspace, and the plan sets how many websites you can connect. See pricing for the limits.
Questions and answers
Content Optimizer questions
What does the Content Optimizer score?
A draft or live page against the pages that rank for its keyword. The score has four parts: topic coverage, length, keyword placement and subheading topics, each counted from the comparison pages' own text.
Do I have to supply the comparison pages?
No. If you leave the URLs empty, MrKeyword uses results it has already observed for the keyword, or its own predicted results, and labels which one it used. You can also add your own URLs.
Does it write the article for me?
No. It shows the topics the ranking pages share, the length range, where the keyword appears and subheading ideas. The writing stays with you.
Why is a topic marked overused?
Your draft uses it far more often per 1,000 words than the comparison pages do. Repeating a phrase more than the pages that rank is unlikely to help.
Can I re-score after editing?
Yes. Fetched pages are cached for 30 minutes, so re-scoring an edited draft does not fetch the comparison pages again.
Is there a limit?
A workspace can run up to 60 analyses an hour. The optimizer is available on every plan.
Sources
- Creating helpful, reliable, people-first content — Google Search Central, on writing for people rather than search engines
- SEO starter guide — Google Search Central, on titles, headings and useful page content
- Influencing title links in Google Search — Google Search Central, on how titles are chosen


