Commentary: AI Pricing Has Not Found a Closed-Loop Model
A ToBeSaaS commentary by Dai Ke argues that AI pricing has not yet produced a fully closed-loop business model. It says outcome-based pricing is difficult to measure, usage-based pricing can produce unstable revenue and invite price pressure, and seat-based subscriptions backed by strong product moats remain a more dependable existing path. The article cites charging experiments by Salesforce Agentforce, HubSpot and Intercom while saying seat subscriptions remain their main revenue source; these claims are presented as the commentary's analysis rather than independently verified figures.
The commentary frames a central commercialization challenge for AI products: technically measurable actions may not represent the business outcomes customers are willing to pay for.

现在全世界都在做AI。大家都在砸钱、造智能体,每天冒出来的Agent比SaaS多得多。可有一个问题,一直没人正面回答:这东西到底怎么收钱?商品要成立,先得能量化成收费标准。能不能做出来,做得好不好,先放一边。你连怎么收钱都没想清楚,生意就没成。这话不中听,但谁也绕不开:AI行业的计价,到今天看,还没有一个模式......
本文来自微信公众号: ToBeSaaS ,作者:戴珂
现在全世界都在做AI。大家都在砸钱、造智能体,每天冒出来的Agent比SaaS多得多。
可有一个问题,一直没人正面回答:这东西到底怎么收钱?
商品要成立,先得能量化成收费标准。能不能做出来,做得好不好,先放一边。