The engagement is not standardised. The scale it is measured against is.

Every site’s visibility problem sits in a different place across the five channels — technical foundation, search architecture, content and intent, entity, AI visibility — so a standardised package would fix the wrong layer for most sites it was sold to. This is diagnosis-led work: the engagement is shaped by what the diagnosis actually finds.

What does not change between engagements is the instrument: the same published scale, the same weights, the same verification discipline described in the methodology.

A strategy built before the diagnosis is a guess with better formatting. Atlas reads the site against the five channels first, and the strategy that follows is the prioritised list of what the diagnosis found — ordered by the published weight of the channel it sits in.

Technical foundation is weighted at 20 of 100 points because its failure is total: everything above it inherits it. That is why prioritisation here is not "everything, eventually" — it is the smallest change that moves the number, done first, measured, then the next one.

Information architecture is the one layer where a mistake compounds rather than staying local — every later template, every later internal link inherits it. Consulting engagements that touch architecture are treated with that asymmetry in mind: get this layer right first, because everything built on top of it is otherwise rework.

Technical SEO, ecommerce SEO, international SEO, AEO and GEO are not five separate vendors’ worth of work. They are five channels of the same visibility system, and a consulting engagement is scoped across whichever of them the diagnosis says are load-bearing for that site — not sold as a bundle of all five regardless of relevance.

Diagnose, map, prioritise, build, measure, compound. Two-week loops rather than a quarterly plan, run by the person who did the diagnosis — not handed to a delivery team once the strategy is signed off.