Errata
Corrections to How to Market to a Machine: What to Do When the Buyer Sends an AI Agent, by Bob Calabritto, IdentiBlock Press. Last updated 2026-10-02.
Corrections to the book are posted here, dated, newest first. Nothing has been corrected yet.
The book is about a subject that moves fast enough to date itself, and chapter 14 is a catalogue of what the author expects to be wrong within a year. When something in the text turns out to be wrong, or a figure is superseded, the correction goes on this page with the date it was made and the printing it applies to. Send one to the contact address in the footer.
Additions
Things the first edition does not cover, either because they were left out or because they happened after the research stopped. The printed text is unchanged.
-
Bing and Google now report AI citations to site owners
- Logged
- Touches
- Chapter 3, section 3.7 (p. 43). Chapter 12, section 12.2 (p. 145). Chapter 13, check 14 (p. 156).
The book tells you to track citation share with a visibility tool or a monthly hour of putting your buyer questions to the assistants. Two platforms now report part of this themselves, free, to any verified site owner. Neither report is mentioned in the first edition.
Bing Webmaster Tools has carried an AI Performance report since February 2026. It shows how often your pages were cited in Copilot and Bing's AI answers, which pages, and the grounding queries behind the citations. The data is sampled. In June 2026 it added intents, topics, citation share and a compare view. It reports no clicks.
Google Search Console added generative AI performance reports on 3 June 2026 and opened them to every site on 31 August 2026. They show impressions in AI Overviews, AI Mode and the AI features in Discover, by page, country, device and date. They report no clicks and no queries.
If you are running the 90-day plan, do this in month one alongside the other instruments: verify your site in each tool, open the report, and write the numbers on your before page with the date. Keep running the twenty questions. Neither report names ChatGPT, Perplexity or Claude among the surfaces it covers, neither shows what the answer said, and neither shows the answers you were left out of.
