Before we touch content or strategy, we check whether the site is even structurally capable of ranking. Here's the exact sequence, in order.
1. Crawlability and indexation, first
Nothing else matters if search engines can't reach or index the page. We review robots.txt for accidental disallow rules, validate the XML sitemap and its submission status in Search Console, and check for stray noindex tags or canonical conflicts left over from a previous build. On larger sites we also look at crawl budget — faceted navigation and parameter URLs can quietly waste it.
2. Core Web Vitals and real-world performance
We measure LCP, INP and CLS on real mobile conditions, not just lab data, because the two frequently disagree. The usual culprits are unoptimised images, render-blocking JavaScript, third-party scripts (chat widgets, analytics, ad tags) loaded synchronously, and an unmanaged web font loading strategy. This isn't purely an SEO exercise — it's a genuine conversion and user-experience factor that happens to also affect ranking.
3. Structured data and metadata
We map the schema.org types relevant to the business — Organization, LocalBusiness, Product, FAQPage, Article, BreadcrumbList — and check that they're implemented correctly, not just present. Title tags and meta descriptions get reviewed against actual search intent for the target query, not treated as a place to stuff keywords.
4. Site architecture and internal linking
We check click-depth from the homepage to every page that matters commercially — pages buried five clicks deep rarely perform, regardless of their content quality. We also look for cannibalisation, where multiple pages compete for the same query and split ranking signal instead of consolidating it, and confirm that pillar pages and supporting content are linked in a coherent topical structure rather than an ad-hoc one.
5. Content gaps and competitive positioning
Search Console query data gets cross-referenced against which pages actually rank for those queries, surfacing the highest-value fix: pages with strong impressions but weak click-through, which usually means a title or meta rewrite rather than a rebuild. We map competitor content against topics the site doesn't cover yet.
The output
A prioritised list ranked by impact against effort — not a two-hundred-row spreadsheet nobody reads. That list becomes the first month of the engagement, and it's the same document we come back to when reporting on what actually moved.