Published addenda

Arguments, read a second time.

Each piece starts with an existing article and adds the context, limitation, correction or alternative interpretation it needs.

Responding to Seth’s Blog

A Minimum Viable Audience Is Not Yet a Viable Market

A sharply defined audience can improve the work, but enthusiasm alone does not prove that the group can sustain its economics or expand by word of mouth.

Responding to MIT Sloan Management Review

AI Search Visibility Is Not the Same as Customer Choice

Brands can become easier for models to retrieve without becoming more desirable, and a recommendation delivered without a visit may be impossible to connect to a real decision.

Responding to Marketing Dive

Brands Cannot Control a Black Box by Adding More Guardrails

AI advertising platforms promise automation while withholding mechanics, leaving marketers to supervise systems whose objectives, attribution, and data flows they cannot independently inspect.

Responding to MarTech

AI-Readiness Frameworks Keep Treating Access as Trust

Eligibility, recommendation, and transaction form a useful sequence, but combining them in one maturity model makes deeper machine access look like inevitable progress.

Responding to KPMG

GEO Cannot Make an AI Trust Your Brand

Schema and knowledge graphs can clarify what a company claims, but technical legibility is not credibility and repeated brand signals are not verification.

Responding to World Economic Forum

Agentic Engine Optimization Is Not the New SEO

Making brand information legible to agents matters, but collapsing search, customer service, personalization, and autonomous transactions into AEO obscures their different risks.