Diagnose, fix, measure, maintain.
Every engagement runs the same sequence, because the sequence is the method. The evidence comes first, the fix comes second, and nothing gets called done until the field data agrees.
Diagnose
Logs and a full crawl before any opinion. The stack is worked top down: can it be crawled, is it being crawled, is it indexed, can it rank, is it being chosen. Each layer is ruled out before moving to the next, so the audit names root causes rather than symptoms.
Typical duration: two to three weeks from access to report.
- A technical audit prioritised by impact and effort, with the log lines, crawl exports and field data attached as evidence
- Template-level Core Web Vitals findings from CrUX, with lab traces for the diagnosis
- A redirect, canonical and indexation map of the site as it actually is
- Structured data validation against both Google's and Schema.org's validators
Fix
Implementation directly on WordPress or static builds, or precise, testable specifications for your development team. Where the CMS is the bottleneck, the fix ships at the edge on Cloudflare Workers instead of waiting for a release.
- Changes made and documented, or a ticket-ready spec with acceptance criteria per item
- Redirect maps generated and validated with Redirect Forge before they go live
- Schema, sitemaps, robots and header changes staged and checked
- Edge rules for headers, redirects, hreflang or bot handling when the platform can't
Measure
A baseline is captured before the first change and the deltas are reported after. Search Console's bulk export goes into BigQuery so the data isn't sampled or expiring; Looker Studio is the client-facing layer, blended with GA4 and Business Profile where they matter.
- A Looker Studio dashboard you own, on your own BigQuery project
- Monthly reporting that separates crawl, indexation, performance and visibility
- Googlebot activity from logs, not estimates
- Plain findings: what moved, what didn't, and what the next fix is
Maintain
Sites regress. Plugins update, booking widgets get injected, a release ships a noindex. A monthly technical retainer keeps a specialist watching the logs and the vitals so regressions are caught in days rather than at the next audit.
- Crawl and log monitoring with alerts on status-code and coverage changes
- Release QA against an SEO checklist before and after deploys
- Core Web Vitals regression tracking per template
- Schema and content-infrastructure upkeep as services and practitioners change
Three ways to work together
Fixed-scope audit
Steps one and three, with a fix specification. A single fee quoted after a short scoping call, a defined delivery date, and a walkthrough of the findings with whoever will implement them.
Technical retainer
All four steps, monthly. For practices and product teams that want the site treated like infrastructure: monitored, maintained and reported on by one person who knows its history.
White-label, for agencies
Design-led agencies with a technical gap subcontract the same work under their own brand through datascimarketing.com. No contact with their clients; findings delivered in their templates if they prefer.
Healthcare practices in South Africa are served through Grow My Practice, which packages the same method with practice-specific content and local search work.
See the brand familyStart with the site you're worried about.
Send a URL and a sentence about what's wrong. A scoping reply comes back with what a diagnosis would cover and what it would cost.