Product Sense Pushup 7: Three Ways to Price AI

Week 7 PSP: Three Ways to Price AI

The three products use different pricing models that reflect how they deliver value to customers.

ChatGPT uses a subscription model where users pay a fixed monthly fee for access to different levels of AI capabilities. This makes costs predictable for users, while OpenAI takes on the risk of heavy users consuming more resources than their subscription covers.

GitHub Copilot primarily uses a per-seat model, where companies pay for each developer using the tool. This makes sense because the value comes from improving individual productivity. However, companies still pay even if some employees barely use it, while GitHub takes on some of the risk of higher usage. Usage limits and additional charges can also shift some costs to customers.

Intercom Fin uses an outcome-based model where customers pay for successfully resolved conversations. I find this particularly interesting because customers are paying for results rather than just access. Intercom takes on more of the risk when the AI fails to resolve an issue, although businesses still have to consider whether those resolutions are accurate and helpful.

For Tada, I think a subscription model similar to ChatGPT would make the most sense initially. Since our product helps PMs turn meeting discussions into actionable tasks and improve team alignment, the value comes from continuous use rather than a single outcome. A per-seat model could also work for larger teams, but it might discourage adoption when our goal is to improve collaboration across different roles.

I think the biggest takeaway is that pricing should reflect how customers experience value, rather than simply how much the AI costs to run.

Avatar

About the author

Leave a Reply