–
Estimate based on the bytes transferred.
Green Web Audit
Verifiable technical data, a CO₂e estimate declared as such, and a roadmap ordered by impact. No invented traffic, no numbers without a source.
–
Estimate based on the bytes transferred.
Actions ordered by estimated technical benefit, not by how easy they are to sell.
Potential detected by Lighthouse
– avoidable bytes identifiedLab and real users are kept separate: they answer different questions.
A single controlled Lighthouse run. Useful for diagnosis, not for describing every user.
CrUX aggregates real-world experience at the 75th percentile, when the public sample is large enough.
Page weight matters even with little traffic: if page loads go up, the impact grows in the same proportion.
The projection will appear after the page has been analysed.
Every number has a type: measured, observed or modelled.
Want to go from report to code?
How to read the audit
Digital emissions do not come from a universal meter. This audit applies the Sustainable Web Design Model v4 to the bytes measured by Lighthouse or, when that test is unavailable, to a limited estimate of the resources declared on the page.
Weight and Lighthouse metrics from the single run; in fallback, the weight is explicitly labelled as a direct estimate.
CrUX Core Web Vitals at p75, only when Google has a large enough public sample.
CO₂e attributed to data centres, network and devices according to SWDM v4.
Free public sources give pageviews that are often incomplete or built on fragile heuristics. The buttons therefore show declared projections — not traffic attributed to your site — and you can replace them with your real page-load figure.
No. The Green Web Foundation verifies the provider's energy declarations. SWDM only reduces the operational data-centre component; network, devices and embodied emissions stay in the model.
It is enough to find bottlenecks. For important decisions it is worth repeating the test, analysing several templates and comparing the results against RUM or Search Console data.
Google Web Risk checks whether the URL appears in known threat lists, such as malware or social engineering. No match does not prove the site is free of vulnerabilities or compromises.