A blog about pricing agentic software — why per-seat pricing is dying, what outcome-based models actually look like in practice, and how to price a product that works while you sleep.
Telcoms already ran the AI-without-outcomes experiment — VMO2 and BT's numbers show the result. SaaS is about to repeat it, unless it prices for outcomes.
AI agents don't sell themselves. Adoption comes down to three things: capability, a credible value exchange, and substitutional spend — not net-new budget.
A value exchange is the credible, believable ROI a customer gets for what they pay. If you can't say it in numbers, you don't have one — you have a vibe.
Every agentic pricing conversation snags on the same question: what are you actually charging for? Here's the distinction that decides everything downstream.
Every seat you sell an agent to replace is a seat you'll have to un-sell later. Here's the math nobody's running yet.
Charging for results instead of access sounds generous. It isn't — it's you underwriting your own product for the first time.
Same old SaaS, same old tricks: per-seat, per-month, per-headache. Agentic products break the assumption pricing was built on.