Now · Work & interests
What's on the table lately.
A few things I'm building, reading and thinking about. I update this when there's something useful to add, rather than pretending every week produces a revelation.
Last updated: 23rd September 2026
Building
I've been developing the SEO + AI Standard: a framework with 10 areas, 32 checks and a scorecard for reviewing a site. It brings together questions that come up in my work and gives me a way to explain how I assess them.
It sits alongside the free website tools, including the AI visibility check, Schema Sniffer and Ask Volume.
I've also been working on repeatable AI visibility measurement. Comparing recorded answers over time helps me see changes that a single score would miss.
Most recent piece: How to tell whether AI is actually recommending your brand.
Reading
A few things I've saved recently. The list updates from my reading feed.
Reading feed refreshed: 2 November 2026.
- How to Win at AEO (Answer Engine Optimization) in 2027: My Top 5 Predictions. Some smart, some I'd maybe push back on (Reddit is still influencing ChatGPT results, it just isn't getting the citations) youtube.com
- Google updates AI Mode and AI Overview with more links, context, carousels, and information on creators. Lip service to a rabid SEO community furious at AI overviews taking away everyone's clicks? Or will adding more obvious clickable links in AI answers send cited sites more traffic? searchengineland.com
- The best-converting homepage button is... 🕹️. Minimal friction = maximum conversions marketingideas.com
- Reddit fell out of ChatGPT's citations, not out of the model — Suganthan. So "being cited" and "being known to the model" are different mechanisms. Who knew suganthan.com
- Google AI Overviews now appear in 39% of U.S. desktop searches. We've long known it was inevitable but here are the numbers just in case there was any doubt. If you're not being cited in AI Overviews you're losing the visibility race searchengineland.com
- Inside 12 months of AI search experiments. When SEO goes south and you still win growthunhinged.com
Thinking about
How much a single AI visibility score can really tell us. The answers can change between runs, so I'm interested in what that variation means and how to describe it usefully.
Also, how to make the use of AI in publishing understandable without making the label do all the explaining. Who takes responsibility for the work seems a more useful question than the label alone.
And why I keep returning to plain text files. They're easy to move, edit and reuse. A fairly modest list of talents, but it turns out to cover quite a lot.
Elsewhere on the site
There's a daily game in the play section and a browser-based music tool called Dog on the Decks. The latter includes a bark button. Neither needs a business case.
I've also been trying different models in the workflows I use. It is useful to see where they help and where they need more checking. It can also become a rather elaborate way of avoiding a smaller task.
A question about your own site?
The free AI visibility check is available if you'd like a small sample of how models respond to questions about your market. If you'd rather discuss a specific problem, you're welcome to contact me.