The journal.
Methodology notes, case studies, and arguments about how brands win in AI answers.
Why ChatGPT doesn’t recommend you (and how to find out in 30 seconds)
Absence, weak positioning, outframing, platform gaps — the measured failure modes behind a missing brand, from 163 scanned domains.
The 67-point gap: when ChatGPT and Claude disagree about you
LastPass scores 73 on ChatGPT and 6 on Claude. The widest cross-platform gaps from 20 measured categories — and why a zero is a suspect, not a finding.
ChatGPT and Claude disagree about you more than you think
Across 795 matched buyer-intent questions, the two platforms disagreed on brand presence 18% of the time. A single score hides that.
Your AI visibility didn’t drop. Or did it? Noise, runs, and receipts
Language models are stochastic. Why one answer is a story, repeated runs are evidence, and alerts should be rare and earned.
Who AI actually recommends for note-taking apps — a measured leaderboard
Notion and Evernote appear in every measured answer; two established products appear in none. A dated, inspectable teardown.
The robots.txt mistake that hides you from AI (a 10-minute fix)
GPTBot, ClaudeBot and CCBot obey a file most teams never re-read after launch. How to audit it without breaking your crawl policy.
How to improve your ChatGPT visibility
A practical guide to making your brand discoverable by ChatGPT and Claude. What actually works, and what is folklore.
AI SEO vs traditional SEO: what's different?
AI search engines work differently than Google. Here is what you need to optimize for if you want ChatGPT to recommend you.
How the presence score works
One number for whether ChatGPT and Claude name you — five runs a week, Wilson intervals, and why we show nothing below twelve samples.