Methodology

Every Crawlixa check is deterministic: the same evidence always produces the same result. Nothing about whether a check passes, warns, or fails is generated or guessed by AI — it's plain, documented, evidence-based code.

Stage 1A runs 28 checks across three categories on a single page you submit. Your overall score is a weighted combination of category scores, with a small number of critical issues (like a page returning a server error, or blocking search engines entirely) capping the score to reflect how much they matter — always shown alongside the uncapped number so nothing is hidden.

Some checks — like word count and internal-link count — are heuristics, not hard rules, and are labeled as such in your report. A result marked "not applicable" means the check genuinely didn't apply to your page (for example, an image-alt-text check on a page with no images) — it's never scored as a failure.