Can you really do outcome-based pricing if you can’t measure outcomes?

Last week I met a Voice AI company. We barely talked product. The real heat was pricing, not “how much,” but “what exactly are we charging for?”

They don’t want per-minute, per-seat, or per-API-call anymore. They want per resolved call, per booking, per qualified lead, per deflection.

Sounds clean. Until you try to define “resolved.”
Who validates it?
What if their CRM says something else?
What if attribution breaks?

At that point, the metric becomes the product. And the infrastructure behind that metric becomes the business model.

Everyone’s going to move from usage pricing to value pricing to outcome pricing.

Can your system measure outcomes accurately enough to build a business on?