When you’re selling AI or usage-based SaaS to the enterprise, the hardest question isn’t how many credits to charge... it’s what those credits are ACTUALLY worth.
At our Monetizing AI Summit, Simon Ooley, CEO of Veles, shared how the best enterprise sales teams are shifting from selling tokens to selling outcomes. His framework reframes pricing conversations around use cases, not units... helping both sellers and CFOs see ROI with confidence.
If you’ve ever struggled to justify a consumption model to an enterprise buyer (cough cough, CFOS, cough cough), this is a must-watch. You’ll learn:
Why credits and tokens have no inherent value until tied to a customer’s outcome
How to structure enterprise deals around use case–based pricing
How this model builds trust, accelerates onboarding, and improves revenue predictability
Why “sell the work, not the token” might be the single most important shift in your sales motion
A credit or token has no inherent value. It only gains value when tied to a customer’s desired outcome. The more a buyer does, the more they get. So how do we price and package when we want customers to do as much as possible? In enterprise sales, especially with a CFO in the room, that is hard. Subscriptions are relationships. We need to get back to delivering value continuously, not chasing a single five-year check.
Simon: I am the CEO of Veles. I am a sales guy, not a pricing theorist. My job is to define and defend the value of AI monetization to customers, especially enterprise buyers. Most of you want to sell to large companies and use consumption or hybrid models. I will cover how to make that work in an enterprise motion.
Traditional good, better, best used features to define value. In AI and modern data tools, packages cannot bound what customers do. A product like Clay can be used in many ways. Predefined bundles either constrain value or leave money on the table. That is why teams turn to consumption. But consumption is tough to justify to a CFO unless you anchor it to outcomes.
Short answer: nothing on its own. A unit only has value when you link it to a customer outcome. Value lives in outcomes: time saved, revenue created, or a defined operational result. Credits gain value when you know how many are needed to achieve that result.
Define deals around use cases, not units. A use case is a scoped outcome with four parts:
Actors: Who is involved.
Workflow or workload: What they will do or what the system will do.
Consumption driver: The meter behind it. Credits, tokens, DBUs, requests.
Success criteria: The measurable outcome.
In discovery, identify the highest value use cases, not just usage levels. Sell the work, not the token.
Price the platform and meter, but present it through use cases.
Tie the requested credit pool to the number and priority of use cases.
In negotiation, trade outcomes, not unit price. If they want to spend less, they get fewer use cases or a reduced scope. This keeps the discussion grounded in business value.
Do not promise “11x ROI” in the abstract. For each use case, show the chain from activity to dollars. Example: automate inbound routing, lift qualified meetings by 20 percent, given a 20 percent win rate and a 50K ASP, expected impact is X. Stack use cases to build a credible ROI story that finance can follow quickly.
Each use case becomes the implementation blueprint. Start with priority one. Deliver that outcome before expanding. Avoid analysis paralysis in flexible AI products by sequencing use cases, not features.
Model as many use cases as possible, then stack rank. Close the top two or three. The rest become a ready backlog for upsell as outcomes are delivered.
Over time you learn how many credits each use case burns and how long it takes to deliver. That makes revenue and usage forecasts more accurate.
Procurement’s job is to buy what the business needs. Use cases help champions justify spend even when budgets span different lines, like shifting from software to labor savings. If there is pushback, prioritize which use cases land now and which later. You are trading scope, not arguing about token price.
Moving to outcome selling is an organizational shift. Best-in-class examples like Snowflake and Databricks center the sales process on use cases and ongoing value. Think in terms of a long-term relationship where the team continuously delivers scoped outcomes tied to consumption, not a one-and-done license event.
Different actions can have different credit costs. Keep types limited to preserve clarity. Start by mapping credit costs to the value of the action and sanity-check against variable cost. Expect to iterate. Pricing is not set in stone for a year. Change it as you learn.
Simon: Credits are just currency. The enterprise buys outcomes. Frame discovery, pricing, negotiation, onboarding, and forecasting around use cases. Sell the work, not the token.