When we audit a company's digital estate, the method matters more than the findings, because findings without a method are opinions with a chart. Ours is built on a rule borrowed from our commercial work: every figure carries its basis, its date, and the honest limits of how it was measured.
The same harness for everyone
Every site in a comparison set is measured by the same instrument, the same day, from the same network. Pages crawled from each site's own sitemap, headers read off the wire, structured data parsed from source, analytics detected from actual network behaviour rather than page claims. If the harness cannot reach a competitor, the audit says BLOCKED. It never says zero.
Ten dimensions
Surface (what is published and indexable), discoverability (titles, descriptions, structured data), measurement (whether the business can even see its visitors), infrastructure (security headers, delivery), speed, content depth and cadence, authority signals, off-page reality, AI citability, and the conversion path. A consumer business and a consultancy weight these differently; the instrument declares its weights instead of hiding them.
Declared out of scope
Some things need tools we may not deploy for a first audit: backlink authority needs a paid index, field performance needs real-user data, brand search needs the owner's console. The audit lists these as out of scope with what it would take to close them. An audit that claims completeness is advertising; one that states its gaps is measurement.
The monthly re-run
The first audit is a baseline. The value is the re-run: the same instrument, the same peer set, movement reported monthly. Improvement becomes a number somebody can be accountable for, which is the only kind of improvement that survives contact with a budget meeting.