Transparent by design
The NXT100 scoring methodology
NXT100 compresses a mobile website diagnostic into five equally weighted signals. The number is useful because the evidence stays visible.
Effective August 20, 2026 · Version 1.0
What the score means
The total score is the sum of five pillars, each worth up to 20 points. A score is a directional diagnostic for the page and moment tested. It is not a ranking guarantee, revenue forecast, legal opinion, or complete accessibility evaluation.
Mobile Lighthouse performance, including loading, responsiveness, and visual stability.
Automated technical SEO signals such as titles, descriptions, crawlability, and descriptive links.
Browser best-practice signals including secure transport, runtime errors, and vulnerable client libraries.
A blended mobile-friction signal using loading, blocking, layout shift, viewport, text, and tap-target audits when available.
Automated accessibility indicators covering names, labels, contrast, language, structure, and alternatives.
Evidence becomes action
Failed or weak audits are ordered by severity. Each finding includes the observed evidence, likely impact, confidence, and one practical next action. When an automated audit does not fully answer the question, NXT100 lowers confidence and recommends human review.
Full versus limited coverage
A full score uses the mobile Lighthouse lab. If that service is unavailable, NXT100 may inspect only the public HTML, response headers, and page structure. That provisional mode is visibly labeled and capped at 66/100 because it cannot measure browser rendering, interaction, layout movement, color contrast, or assistive-technology behavior.
Score bands
- 80–100: strong automated foundation; protect the score and inspect the remaining edge cases.
- 60–79: workable foundation with visible opportunities.
- 0–59: material technical or usability friction needs attention.
Source and variance
The full engine uses the Google PageSpeed Insights API and Lighthouse lab data with a mobile profile. Network conditions, page changes, third-party scripts, test infrastructure, and measurement variance can change results between runs. Use trends and repeated evidence—not a single number—as the decision standard.
Responsible use
Submit only public URLs you are authorized to evaluate. NXT100 does not scan private networks, login areas, or protected content. If a result could affect legal, security, or accessibility decisions, use a qualified human reviewer.