Latest articles
Current guidance on AI search retrieval, measurement, crawler control, agent-ready websites, and commerce, with sources, methods, and limits.
Current research and practical guidance
UCP and ACP are not ranking switches. Here is what they change, what they do not, and why product data, live inventory, and checkout have to agree.
WebMCP gives browser agents structured tools, but reliable tasks still depend on semantic HTML, narrow actions, strict validation, consent, and clear receipts.
Search, training, and user-requested retrieval are different jobs. This guide maps the major AI crawlers to the controls and checks that actually apply.
There is no honest universal AI rank. Measure citations, repeated prompts, referrals, conversions, and revenue separately, with every limit visible.
AI search can split one question into many searches and build an answer from changing sources. Here is how to make the right facts easy to retrieve and cite.
A citation, crawler hit, visit, and sale prove different things. This is how I keep AI search sources, tests, interpretation, and results honest.
The parts of AI SEO I keep watching
How answers get built
Follow the path from the original question to follow-up searches, candidate pages, grounding, citation, and the final answer.
What the numbers mean
Keep visibility, citations, visits, conversions, and revenue separate so the report can support a real decision.
Can an agent use it
Check whether crawlers and browser agents can reach the site, understand the interface, and complete a task safely.
From discovery to purchase
Keep product facts aligned across pages, structured data, feeds, live inventory, checkout, and new commerce protocols.
Show the work
Publish the sample, conditions, method, uncertainty, and limits so a reader can judge the finding instead of trusting a score.
What changed
Track platform rules, reports, crawlers, standards, and claims that need a closer look when the system changes.