Search, engineered from the logs up.
YSM Digital is a one-person technical SEO studio in Johannesburg. Crawl, indexation, Core Web Vitals and structured data, diagnosed from server logs and field data rather than opinion, for healthcare practices, agencies and product teams.
- location
- Johannesburg, ZA
- coordinates
- 26.20 S, 28.03 E
- timezone
- SAST, UTC+2
- experience
- 7+ years
- capacity
- 1 senior specialist
- pipeline
- logs → BigQuery → Looker Studio
Most SEO advice starts at the content layer. Most ranking problems don't.
A page that isn't crawled can't be indexed; a page that isn't indexed can't rank; a page that ranks slowly loses to one that ranks fast. The studio works that stack in order, and it never recommends a content fix for what the logs show is a crawl problem.
What the studio does
Crawl and indexation
Raw server and CDN logs, not sampled crawl stats. Googlebot hit frequency verified by reverse DNS, status codes over time, crawl budget by URL template, canonical resolution, noindex leaks in headers and meta.
Core Web Vitals
Field data decides; lab data diagnoses. Work at the template level: hero images and render-blocking plugins for LCP, injected booking widgets for CLS, tag managers for INP.
Structured data and AI search
JSON-LD that validates in both the Rich Results Test and the Schema.org validator. Entity clarity, consistent practitioner credentials and quotable declarative copy so AI Overviews, ChatGPT and Perplexity can cite the right source.
Edge SEO
When the CMS is locked or the dev queue is long, Cloudflare Workers ship the fix: header injection, redirect logic, bot response shaping, hreflang.
Measurement
Search Console bulk export into BigQuery, reported in Looker Studio. GA4 and Business Profile blended for the client view; log-level detail kept in internal views.
Tooling
Python where the work is repetitive: a bulk redirect mapper, an audit pipeline, scrapers and converters. Working code, not pseudocode, handed over with the findings.
Real technical work, with outcomes
Notes from the work
Server logs are the only crawl data that isn't sampled
Search Console shows a sample. Logs show every request. The standard cuts, and how to verify Googlebot before you trust a line.
2026-08Field data decides, lab data diagnoses
Why a green Lighthouse score can still fail Core Web Vitals, and the four usual suspects on WordPress healthcare sites.
2026-07Structured data for practices, and what AI engines actually read
MedicalClinic and Physician JSON-LD, two validators that disagree, and the entity work that gets a practice cited.
Three pieces at launch. The blog is where the method gets written down, so clients can read the reasoning behind a recommendation instead of taking it on faith.
All writingSend the site. Get a diagnosis, not a pitch deck.
A short scoping call, then a fixed-fee audit or a monthly technical retainer. Replies from Johannesburg, usually the same working day.