How to price when AI keeps collapsing delivery costs: the deflation curve (Auguste), value-based anchors (Edwards, Koerner), productized tiers (Aristide), and Hormozi's offer equation applied to AI services. Ends with a packaging worksheet.
AI keeps collapsing what it costs to deliver expert work. Frédéric Auguste put the math in public: a heavy agentic day burns $30–60 of list-price compute, so a $25,000-style audit can go on sale for $299 and still make money. That is deflation, and it is coming for every service priced on hours. This paper is about how to price anyway. The anchors come from people actually selling AI services today. Aaron Edwards resells AI agent services to local businesses at $2,000 setup and $500 a month — after hitting $1M ARR with DocsBot AI in February 2026. Chris Koerner trains consultants to charge $500–$5,000 a month for AI implementation. Adam Erhart runs a $10,000-a-month one-person agency on AI and HighLevel. Marc Aristide sells a single-feature SaaS at $97 a month, with roughly a third of his revenue coming from rebilling.
Summary drawn from the paper's executive summary (Human-N-Loop research series, October 2026 draft).
internal strategy doc first, client-facing second; informs every CED price Chris sets.
Sources named in the research: Auguste, Edwards, Koerner, Aristide, Hormozi, Adam Erhart.