Pricing AI Is Hard: 8 Lessons (and a Few Scars) from 12 Years at Salesforce
SaaS pricing wasn't built for AI. Join Glenn Dearing & Jordan Zamir to learn why predictable beats perfect, how to use fungible credits, and more.
%20%20from%2012%20Years%20at%20Salesforce%20(1).png)
%20%20from%2012%20Years%20at%20Salesforce%20(1).png)
Pricing AI is hard in a way SaaS pricing never was: the costs are real, they move week to week, and the old playbook of seats and platform tiers wasn't built for it. Glenn Dearing spent 12 years running pricing at Salesforce and later led it at UKG. He's joining Turnstile’s CEO, Jordan Zamir, to talk through what that taught him and how much of it survives the shift to AI.
Glenn is bringing eight hard-won lessons, each with the story and the scars behind it.
A few you'll hear:
- Predictable beats perfect. A pricing metric customers can't forecast is worthless — no matter how well it tracks value.
- Fungible credits. How to pass falling model costs on to customers without triggering a painful repricing event.
- Ship before you're right. We shipped a token model we knew wouldn't last, because a quarter of real feedback beats a quarter of being right on paper.
- Legibility over transparency. Publishing your price isn't enough until customers can explain your model back to you.
He'll test every lesson against the AI era — what still works, what GenAI just broke, and what revenue and finance leaders should be building now. Jordan joins to dig into the infrastructure that has to keep up: the metering, quoting, and billing underneath pricing that changes.



























Close deals. Get paid. Know your numbers.
Turnstile connects quoting, billing, and financial reporting in one place — built for the complexity of how sales-led startups actually sell.

