Content readiness
Every article is graded against these criteria. This page is the rubric itself — the same definitions the grader uses.
What this score is, and is not
It is a content readiness indicator produced by our own checks. It is not a score from any search engine, it does not predict rankings, and passing these checks does not prove an article is factually correct. Articles scoring below 70 are held for your review.
Judged by reading the article
These need comprehension, so they are assessed by a language model in a single pass against the criteria below.
| Criterion | What it means | Points |
|---|---|---|
| Search intent | The article answers the intent behind the topic, and answers it early. | 12 |
| Topic coverage | The article covers what a reader on this topic needs, without padding. | 12 |
| Specificity and usefulness | Concrete, actionable detail rather than generic statements. | 12 |
| Unsupported claims | Every factual claim is either common knowledge or supported by a cited source. | 14 |
| Title | The title describes the article accurately and reads naturally. | 8 |
| Meta description | A description that summarises the article and would earn a click honestly. | 6 |
| Readability | Clear sentences and paragraphs appropriate to the audience. | 8 |
| Heading structure | Headings describe their sections and form a sensible outline. | 6 |
| Topic cluster coverage | The article covers the supporting concepts a reader expects alongside the primary topic. | 8 |
| Brand voice | Matches the customer's stated voice and spelling convention. | 5 |
Measured in code
These are counted exactly rather than estimated. Word limits in particular are verified by code — never taken from the writer's own claim about its output.
| Criterion | What it means | Points |
|---|---|---|
| Primary keyword placement | The primary keyword appears where it carries weight: the title, the opening, and at least one subheading — read naturally in each. | 8 |
| Natural keyword use | Keywords are not over-repeated. | 8 |
| Length target | Within the customer's configured word range. | 5 |
| Heading hierarchy | No skipped levels; exactly one H1 (the post title). | 4 |
| Internal links | Within the configured range, pointing at verified eligible pages. | 4 |
| Image alt text | Every image has descriptive alt text. | 2 |
| Description length | Between 70 and 160 characters. | 2 |
Total available: 124 points
Can an article score 100?
Yes — and only by being genuinely good, which is the point. Two of the criteria above pull in opposite directions:
- Primary keyword placement rewards the keyword appearing in the title, the opening and a subheading.
- Natural keyword use starts at full marks and removes points as repetition climbs.
So full marks on both requires natural placement at a sane frequency. An article stuffed to satisfy the first loses more on the second than it gains, and its total falls. There is no keyword density target to hit, because a target is something to game.
What happens to a low score
An article below 70 is held for your review rather than published. So is any article containing a claim we could not match to a source, and any article where a check could not be completed — because in that case we do not actually know.
We do not regenerate repeatedly to chase a higher number. That would spend your balance on our own second opinions.