Frameworks, core principles and top case studies for SaaS pricing, learnt and refined over 28+ years of SaaS-monetization experience.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
.png)
Once your packages map to your segments, the next decision is the pricing metric. For Agentic AI products, this may be the most important decision in the whole exercise. Most of the noise about "outcome-based pricing," "consumption pricing," or "seat pricing" comes down to one question: what do you measure and bill for?
Think of your options as a range from fixed to variable:
Fixed and predictable (left side): named-user pricing, perpetual on-prem licenses, flat annual fees.
Variable and usage-based (right side): storage, tokens, events, and other infrastructure-style usage.
The goal is to find the right spot on that range for your company and your market, not to chase whatever metric is trendy.
There is no single "correct" metric. Every choice balances what the customer will accept against what your economics need. Three forces pull from the customer side and four from the business side.
Customer side
Value alignment. The metric should track the value the customer actually gets. A metric customers can tie to their own success is a win for both sides.
Risk perception. If buyers see your product as risky or unproven, they often prefer a variable metric that limits their downside: they pay only for what they use. A confident buyer will accept a more predictable, committed metric.
Mental anchors. Find out how customers are used to buying this kind of product and what meter they expect. If they are not open to a new meter, innovating here may not pay off.
Business side
Cost of goods sold. For inference-heavy AI products, a heavy user on a per-seat plan can cost you more than you charge. That can push you toward credits or tokens, even when those meters do not match the value the customer gets.
Consumption fit. Some software earns its value just by being available, like a car that sits in the garage but is there when you need it. An ERP that runs quietly in the background is similar. Moving products like these to usage or consumption pricing may not pay off. Know what kind of product you have.
Competitive action. If a large player like Amazon or Google enters your space with a certain metric, you may have to follow, even when it is not ideal for you.
Implementability. The sharpest, most exact metric is useless if you cannot run it. Can you meter it? Send it to billing? Will customers understand it? Can you defend it when they push back? Outcome metrics like "resolution" must be clearly and objectively defined, and you have to accept that some customers will argue about them. The bet is that the clarity and value alignment are worth the friction.
For Agentic AI, we built a smaller framework at Monetizely, one we call the AMS (Agentic Monetization Spectrum). It helps you find the most important criteria, or the tie-breakers, for a quicker read on which pricing metric to use, instead of working through the full seven criteria above.
Join companies like Zoom, DocuSign, and Twilio using our systematic pricing approach to increase revenue by 12-40% year-over-year.