Services
SEO Services Social Media Website Development AI Automation About Reviews Portfolio Blog Contact Free SEO Audit
Author
Founder & Digital Strategist
I have worked in search since 2014, and set this up as a one-person operation in 2018. I write here about work I have actually done, and every article says what its numbers are built on.
Technical audits, content structure and the measurement behind both.
Read more →Being cited by AI Overviews and assistants, and reporting whether it happened.
Read more →Fast, structured sites built so the SEO work has somewhere to land.
Read more →Strategy and content that supports search rather than running beside it.
Read more →Most agency writing quotes figures with no source. I read eleven agency blogs while building this one and not a single post said where its numbers came from. So every article on this site carries a label saying what it is built on, and names the service page it supports.
The one I am most willing to be checked on: before rebuilding this site I audited all 24 pages of the old one and exported three months of Search Console data. Not one of the 24 had a meta description, so Google was writing every snippet across 50,541 impressions itself. One query sat at position 9.35 — page one — and produced 558 impressions and zero clicks. The URLs, the export date and the method are published, not summarised.
These are the planned articles — the titles and angles are real, the dates and read-times are placeholders until each one is written.
Our own measured data
50,541impressions audited. Zero meta descriptions.
Before rebuilding this site we read all 24 pages of the old one and exported three months of Search Console data. Every single page was missing its meta description, so Google was writing all 50,541 snippets itself. This is the full working — the URLs, the export date and the method.
Hands-on analysis
0buttons in Search Console that report this
Share of voice in AI Overviews is a number almost nobody reports, and there is no button in Search Console for it. Here is the manual method, what it can and cannot tell you, and why the sample size matters more than the score.
Google documentation
1test that decides whether a city page should exist
Most multi-location pages are the same page with the city name swapped, which is the textbook definition of a doorway page. The test we apply before writing one: if swapping the city name leaves the page still true, it should not exist.
Our own measured data
0clicks from 558 impressions at position 9.35 — page one
One query on this site sits at position 9.35 — page one — and drew 558 impressions and not a single click in three months. Ranking was never the constraint. What Search Console is actually counting explains it, and it changes what you fix first.
Google documentation
2fixes that account for most of the gain
INP replaced FID two years ago and a lot of advice never caught up. What Google actually documents as a ranking signal, what is only a user-experience win, and the two fixes that account for most of the gain.
Google documentation
1word the policy turns on: purpose
Not "AI content is penalised" and not "AI content is fine" — the policy turns on purpose, and the wording is narrower than either camp claims. The relevant sentences, quoted, and what they mean for a small team publishing weekly.
Our own measured data
7→4workflows automated, then kept
We automated seven internal workflows and kept four. What the three failures had in common was that the checking took longer than the doing. An honest account of both halves.
Hands-on analysis
10questions — answered about ourselves
Ten questions that are uncomfortable to answer if the work is not real: who owns the repository, what happens to the site if we stop paying, and which of these screenshots is a live URL. We answer all ten about ourselves.
Hands-on analysis
3connections that actually hold up
Social links are nofollowed and social shares are not a ranking factor, both of which are true and neither of which means the channels are unrelated. The three connections that do hold up, and how we measure them.
Free 30-minute audit · No obligation · Response within 24 hours
Analytics cookies
We would like to measure which pages get read, using Google Tag Manager. Nothing loads until you choose, and none of it is used for advertising. What we collect →
