This is the least glamorous claim in the series, and one of the most expensive.
The CIM says the pricing model supports the growth plan: flexible packaging, room to expand through modules and tiers, and usage-based pricing for AI on the roadmap.
Then I ask to see the billing and rating process, start to finish, for one billing cycle. More often than not, it's a set of reports. Revenue operations and finance pull usage and entitlement data out of the product, reconcile it in spreadsheets, and create invoices by hand in the ERP or billing tool. Every legacy price book and every custom deal adds another rule someone has to remember. There's no system. There's a process that two or three people know how to run.
That creates three problems at once.
First, it caps pricing and packaging options. If the product can't meter usage and the billing process can't rate it, the company can't move to consumption, outcome or AI-based pricing without a major project. The monetization levers in the value creation plan are locked until someone builds the plumbing.
Second, it's a large manual effort full of error and audit risk. Manual rating produces errors. Errors produce credits, disputes and findings, both from external auditors and from customers who exercise their audit rights.
Third, it almost always leaks revenue. Overages go unbilled. Price escalators get missed at renewal. Customers use entitlements they never paid for. Discounts outlive the contract terms. In a manual process, the mistakes run in the customer's favor more often than not.
AI makes all of this more urgent. AI is usually priced on consumption, and when a customer pays for usage, they expect to see that usage, and what it's going to cost, inside the product, as it happens. A spreadsheet at month-end doesn't meet that expectation. Neither does a surprise invoice.
The questions I ask:
Show me every pricing model you have, with customer count and ARR on each.
How many legacy pricing models and price books are still active? Show me the contracts that go with them.
How many contracts carry non-standard terms, custom rates, caps, escalators or credits, and how are they tracked?
Does the product meter usage and enforce entitlements, or does finance build invoices from a data extract?
Can a customer see their usage, spend and invoices in the product's admin area?
How many people, and how many days, does each billing cycle take?
How many invoices were disputed, credited or reissued last year, and why?
How long would it take to launch usage-based pricing for an AI feature?
The single most useful piece of evidence is a reconciliation. Pick 20 customers yourself. Compare what the product says they used and are entitled to against what they were actually invoiced. It will tell you more about leakage than any interview.
For portfolio companies, I'd put a revenue leakage review in the first hundred days of any new hold, and well before any exit. It often pays for itself. Then connect metering, entitlements, rating and invoicing into one flow, and give customers a usage and billing view in the product. That work unlocks every pricing move in the value creation plan, and it's the foundation any serious AI monetization story needs.
The red flag: billing is a monthly spreadsheet process owned by two people in finance, and the growth plan assumes new pricing or AI consumption revenue. The pricing upside in the model isn't available until the infrastructure exists. Budget for it.
Next week, the last claim: pricing, discounting and CPQ.
