Intended Consequences

The episode contains several useful, non-obvious insights about AI's role in brand discovery and website optimization (AI as 'third partner,' training AI through content specificity, shift from article ranking to sales page optimization), but these core ideas are repeated and explored incrementally rather than densely packed. The conversation meanders with mutual affirmations and hypotheticals, diluting insight density.

“we need to, as brands train AI to be sales reps for our companies by giving it enough detail to help the AI user. Decide whether or not we're a good option”

“The average length of a Google query is 4.2 words. The average length of a prompt is 23 words.”

The 'AI as third partner' framing is novel and useful, and Crestodina's point about websites as knowledge sources for AI training is relatively fresh thinking. However, the underlying advice (clarity, specificity, focus on bottom-funnel conversion) is fairly standard in digital marketing. The episode doesn't challenge conventional wisdom sharply or present truly contrarian arguments.

“AI is gonna be the third. The third partner in every relationship”

“Each of us has the opportunity to plug holes or to remedy gaps and misconceptions in AI training data. By simply updating our own content”

Andy Crestodina is a well-known digital strategist with genuine expertise in SEO and conversion optimization, and Brian Massey is a conversion optimization veteran with decades of practitioner experience. Both are operators, not pure thought-leaders. However, this is a peer conversation rather than an interview of someone at extreme scale or with novel institutional credentials.

“I am a dual threat marketer or hope to be”

“a man who's dedicated decades of his life to understanding why people do and don't take action within a digital experience”

The episode includes some concrete data (SEMrush research on query vs. prompt length: 4.2 words vs. 23 words; Crestodina's own click-through rate decline from search results), but much of the discussion is speculative about future AI behavior ('I think that', 'I assume', hypothetical scenarios). Few named companies, specific client examples, or dollar figures are provided. The evidence is mostly anecdotal and forward-looking rather than grounded in demonstrated results.

“The average length of a Google query is 4.2 words. The average length of a prompt is 23 words.”

“I've gotten a couple of, AI recommended use, which is, it's telling. So I'm glad that I've been blogging since 2006”

Brian Massey asks reasonable clarifying questions and pushes back occasionally (e.g., on whether AI users really have long conversations, on the future of agentic AI). However, many of Crestodina's claims go unchallenged ('I don't have any data that would bear that out' is Massey's honest response, but he doesn't press further). The conversation is collegial and thoughtful but lacks sharp follow-ups or productive friction. There's substantial mutual agreement and few moments of genuine disagreement.

“I don't have any data that would bear that out. So the answer is, I don't know.”

“So there's a new importance of websites, which is that they are feeding AI and that it is your best chance to tell your full story”