The reason to audit your website environmental impact is rarely environmental conviction. It is usually a form.
A tender asking about your digital carbon footprint. A client with reporting obligations passing them down the chain. A board paper that needs a number in it. Whatever prompted it, you now need a figure you can defend rather than one you found on a free tool.
Auditing website environmental impact properly takes an afternoon. Doing it in a way that survives questioning takes the same afternoon plus a document.
Website Environmental Impact: A Practical Audit Guide
Not a measurement. A documented estimate with a stated method.
That distinction is the whole difference between an audit that helps you and one that creates exposure. Anyone competent assessing your response will ask which tool, what conditions and what scope. Having those answers ready is more valuable than a favourable number.
Step one: decide the scope, and write it down
Website delivery only, in almost every case. The energy associated with serving your pages, transferring them and processing them on visitors’ devices.
Not your office electricity. Not your staff laptops. Not your cloud services generally. Those belong in a wider assessment and mixing them in produces a figure nobody can interpret.
State the boundary explicitly in whatever you publish. A figure without a scope is not a figure.
Step two: choose representative pages
Your five to ten most-visited, taken from analytics rather than assumption.
The homepage is rarely representative. Product pages, galleries and anything a marketing campaign built are usually heavier and frequently busier.
Step three: measure what is measurable
Page weight and request count from your browser’s network panel, cache disabled.
Core Web Vitals under the same conditions.
Record the conditions every time: which device, which connection, first or repeat visit, and the date. Skip this and nothing you measure later is comparable, including your own follow-up audit.
Step four: estimate what is not measurable
Run a carbon calculator on the same pages. Note which tool and which version.
These take your page weight and apply modelled assumptions about energy intensity, network transfer and hosting. Two tools disagree about the same page because their assumptions differ, which is a property of modelling rather than a fault.
Use one consistently. The trend matters more than the absolute figure.
Step five: check hosting honestly
The Green Web Foundation’s hosting check tells you whether your provider appears in its directory of hosts claiming renewable power.
That is a claim by the host, not an independent audit. If hosting features in anything you publish, ask the provider what the claim actually covers.
Hosting is one portion of the picture. Network transfer and visitor devices are governed by what your pages send, which is the part you control directly.
Step six: multiply by traffic
A heavy page nobody visits matters less than a moderate page with substantial traffic.
Combining page weight with monthly views tells you where effort actually returns something, and it frequently reorders the priority list from what the page weights alone suggested.
Step seven: prioritise and act
Images first: display dimensions, modern formats, lazy loading below the fold. Largest single win on most sites.
Then unaccounted third-party scripts, then fonts, then JavaScript.
Re-measure after each change under the same conditions. Where performance and environmental impact both apply, treat them as the same lever: reducing unnecessary data transfer and processing supports a faster experience while lowering estimated digital carbon, though environmental figures remain modelled rather than directly measured.
Reporting it without creating a problem
Environmental claims must be accurate and substantiated. The Competition and Markets Authority’s Green Claims Code sets out the standard.
Report the figure, the tool, the conditions, the date and the scope. Describe it as a modelled estimate covering website delivery.
Avoid neutrality claims, which need evidence a website audit cannot provide. A specific measured reduction with a stated method reads as more competent and carries no exposure.
Then diarise the next audit. Quarterly against the baseline catches drift before it becomes a project.
Our website sustainability audits follow this method and hand over the documentation, which is the part that makes it usable in a tender.