Method
How the section check works.
A semantic content optimizer is only worth trusting if you can see what it did. This page is the whole method, including the parts that are still missing.
The input is deliberately small
One target query and one draft. No project setup, no crawl, no keyword list to import. Semantic Checker assumes you already chose the query — from Search Console, a rank tracker, or a client brief — and that what you need now is a second opinion on the draft.
You can paste the text or give it a published URL. Importing a URL is free and never costs a credit, so a failed fetch costs you nothing.
A draft becomes a map
Sections are split at your headings. Before anything is scored, the parser drops the furniture that is not really your article — breadcrumb preambles and trailing Related posts or Keep reading blocks — so recommendation widgets cannot drag a verdict down.
How a section gets its verdict
Each section is compared with the target query and given an alignment figure, which falls into one of four bands:
- —Aligned— 75 and above. The section’s wording lines up with the query.
- —Mixed — 50 to 74. Related, but the emphasis has moved. Worth a read.
- —Off-topic — below 50. Little direct query wording, and the dominant terms belong somewhere else.
- —Mixed — short or context-dependent sections get a limited-evidence note and a review direction instead of being hidden behind an Unknown result.
Every flagged section carries the excerpt that produced the flag and the terms the section is actually dominated by. The claim is always checkable against your own text.
Why the score is only a summary
A single number turns editing into a game with a win condition. The well-documented result is writers stuffing suggested terms into sentences that did not want them, until the meter turns green and the prose reads worse than the draft they started from.
A report here shows the summary score, its component signals, each section verdict, and the evidence. The score organizes the edit; it is not a ranking prediction or a target to game.
What it cannot tell you
This is an early product and the honest description matters more than the flattering one.
- —Not every result page can be read. A SERP coverage score appears only when at least five organic competitor pages yield enough text. Otherwise the local report is saved and the credit is refunded.
- —The comparison is lexical, not an understanding model. It measures recurring topic language across readable pages. A section that covers the same idea with entirely different wording can still be under-scored, so every verdict includes evidence.
- —The current scores are lexical diagnostics.They summarize wording and section consistency; they do not claim to reproduce Google’s ranking systems.
- —English only, one query at a time. Analysis assumes a US English context.
- —No rerun comparison. You can analyse an edited draft, but the tool will not diff it against the previous report for you yet.
The upside of that narrowness is real: no language model sees your draft, and each live source used by the report remains visible.
Try it on something you have published
The fastest way to judge any of this is on an article whose weak spot you already suspect. Eligible new users get 2 analyses and no card is required.
If you want to inspect the output first, open the public example report. It keeps the score breakdown, section evidence and limitations visible without exposing a user report.