DATA SOURCE
Page-level, mobile lab evidence.
The primary audit uses Google PageSpeed Insights and its Lighthouse result for the submitted URL. The scan records the returned category scores, diagnostics, page metadata, and available lab metrics at a specific point in time.
Lab results can vary between runs because network, server, and test conditions vary. Real-user field data should be preferred when it is available.
SCORE CALCULATION
A composite indicator—not a business forecast.
The overall health score begins with the available Lighthouse category scores:
The current implementation then subtracts two points for each normalized audit issue and constrains the result to the product’s supported score range. This makes the overall score an internal prioritization aid; the original category scores remain the primary evidence.
AUDIT SCOPE
Measured and not assessed.
Measured when available
One submitted public URL using a mobile PageSpeed Insights audit
Lighthouse Performance, SEO, and Best Practices category scores when returned
Lab-based loading and stability signals available in the audit response
Page title, meta description, sampled images, and selected HTML SEO signals
Diagnostics returned by the audit and page-level checks
Not assessed
—Revenue, lead volume, conversion rate, or commercial loss
—Guaranteed rankings, traffic, or search-result appearance
—A complete crawl of every page on the website
—Manual accessibility testing with keyboards or assistive technology
—Penetration testing, legal compliance, or a complete security review
—Content accuracy, brand suitability, or customer sentiment
RECOMMENDATIONS
Evidence, interpretation, and action are separate.
Each report finding should distinguish the observed audit evidence from our explanation and the recommended action. References point to published guidance where applicable. A recommendation describes a possible technical improvement; it does not guarantee a commercial or search outcome.
CONFIDENCE
Confidence follows the evidence.
Direct Lighthouse failures and captured page metadata have higher confidence than inferred opportunities. Features that were not measured are labeled “Not assessed.” Optional business tools—such as chatbot, QR, local SEO, or forms—should only be recommended after their relevance is established.
The report links back to this methodology so recipients can verify how conclusions were reached.