How Quillly Scores Your Content: The SEO Score Explained

turned on monitoring screen
Photo by Stephen Dawson on Unsplash

Every time your AI creates or updates content, Quillly returns an SEO score from 0 to 100 with the saved item. It's a single, honest signal of whether a draft is ready to rank, designed to be acted on rather than admired — and it's the same pipeline scoring this page. As of August 2026, it's also what's behind quillly.com's own blog: 133 published posts averaging 87/100.

What the SEO Score Measures#

The score is a weighted average of 14+ categories covering structure, metadata, keywords, links, and media. The heaviest categories carry the most points, so fixing them first moves the number most:

Quillly SEO score category weights, with keyword optimization and meta tags carrying the most weight

The Categories, in Plain Terms#

Every category checks something a searcher or a search engine actually cares about, not an arbitrary rule:

Table

Category

What it actually checks

Keyword Optimization

Does the primary keyword appear in the title, opening paragraph, a heading, and the body — naturally, not stuffed?

Meta Tags

Are meta_title and meta_description present, correctly sized, and do they include the primary keyword?

Heading Structure

Is the page organized into ##/### sections a reader (and a crawler) can skim?

Internal Linking

Does the page link to 2-3+ related pages on the same site, with varied anchor text?

Readability

Are sentences short enough and the voice active enough to read at an 8th-9th grade level?

Content Length

Is there enough substance to fully answer the query — density, not padding? See how long a blog post should be.

External Sources

Do claims point to real citations instead of asserting numbers with nothing behind them?

E-E-A-T Signals

Are there first-hand expertise signals, data, and freshness cues a search engine's quality systems look for?

These map closely to what Google's own SEO Starter Guide and guide to helpful, reliable, people-first content describe as the fundamentals of a rankable page — Quillly's score is a machine-checkable proxy for those same signals, not a separate standard.

Key Rules to Know#

A few concrete thresholds drive most of the score:

  • Meta tags — give every item a meta_title of 50-60 characters and a meta_description of 120-155 (keep it under 160). Include your primary keyword in both.

  • Internal links — add at least 2-3 links to related content on your site, using ai-driven internal linking rather than the same anchor text every time; identical anchors are flagged.

  • Visual density — aim for one visual every ~400 words, spread across sections. Below roughly one per 500 words you lose points, and any 700-word stretch with no image is flagged. Photos count, and so do code-rendered visuals — a Mermaid diagram, a data chart, or a custom HTML layout, each created with an image: code fence and each needing alt text.

  • Structure and readability — use ## and ### headings, short paragraphs, and transition words; do not repeat the title or a byline in the body, since the page chrome already renders them.

  • Keywords — pick a primary keyword you can realistically win. Our AI keyword research playbook covers how to check a keyword's real difficulty before you commit a draft to it. Once the keyword is settled, seven examples of a blog post compares the formats it could take — each with the impressions and search position it earned here.

Run an on-page audit against these same categories on any existing post to see where your current content already stands.

How Your Agent Should React#

Do not guess at fixes. The loop is designed to be mechanical:

Flowchart of the improve-check-improve loop: create draft, get SEO patches, apply them, recheck score

After a draft is saved, have your AI call get_blog_seo_patches. It returns a prioritized list of changes with the projected point gain for each — for example, "+8 if you fix Meta Tags" — plus concrete per-category suggestions. Your AI writes the actual edits (the tool gives suggestions, not literal find/replace text) and applies them all in a single update_content call using patches, then re-checks with check_blog_seo. This tight improve-check-improve loop is the fastest way to a high score, and it's the same loop that produced the 87 average mentioned above.

What Is a Good Target#

70 is the publishing floor — Quillly's pre-flight guidance is to reach a score of at least 70, with valid meta lengths, before publish_content. Some websites set a stricter min_publish_score in their content policy, and publish_content is refused below whatever that site's floor is.

Aim higher than the floor, though: 90+ earns an A, and it's achievable — most posts published through Quillly land in the high 80s and 90s, and the score caps deliberately (a page can't hit 90+ on thin content or missing sources, no matter how well-optimized the rest is) so a high score can't hide a weak page. If a draft is below 70, run the patches flow before publishing rather than shipping a weak post.

Key Takeaways#

  • The SEO score is a weighted average of 14+ categories, not one number pulled from thin air — the heaviest ones (Keyword Optimization, Meta Tags, Heading Structure) move the score the most.

  • 70 is the publishing floor; 90+ is the real target, and it's capped deliberately so thin or unsourced content can't fake its way there.

  • The fix loop is always the same three calls: get_blog_seo_patches, one update_content, then check_blog_seo — repeat until the score clears your target.

  • This isn't theoretical: it's the exact scoring pipeline behind quillly.com's own 133 published posts, averaging 87/100.

Frequently Asked Questions#

What's a good SEO score for a new post?#

Aim for 85+ before publishing if your website's policy allows it — most posts published through Quillly land in the high 80s to mid-90s once the patches loop runs once or twice. Below 70, treat it as a draft, not a ready post.

Why does my score stay capped even after fixing several categories?#

The score caps deliberately when a critical category still fails — for example, thin content or zero external sources — even if every other category is strong. Check criticalIssues in the get_blog_seo_patches response; those are what's holding the cap down.

Does the SEO score guarantee a page will rank?#

No. It's a proxy for on-page fundamentals, not a ranking guarantee. Indexing status, domain authority, and SERP competition still apply after a page scores well.

We've watched this directly on our own blog: a post can score 95 and still sit unindexed for weeks on a low-authority site, because indexing and ranking are separate problems from on-page quality. The score tells you the page is ready; it doesn't control what a search engine does with it next.

How often should I re-run check_blog_seo?#

Every time you materially change a post's content, keywords, or metadata. Minor edits like fixing a typo don't need a re-check; adding a section, changing the primary keyword, or rewriting the intro does.

Can I publish below the minimum score anyway?#

Only by passing force: true to publish_content, and only when you explicitly intend to — the normal path is get_blog_seo_patches → one update_content call → check_blog_seo until the score clears.

Ready to try it? Connect your AI and ask it to write, score, and publish a post — or start from Getting started with Quillly.

Share