Start with the value event
Choose the model that charges closest to the moment the customer receives measurable value. This principle connects the price to what the product delivers rather than to a familiar format such as a subscription.
Compare what the customer pays for
- A one-time sale fits a fixed result with little ongoing service.
- A subscription fits repeated value that would be missed after cancellation.
- Usage-based pricing fits value and variable cost that both scale with consumption.
- A transaction fee fits a product that enables and protects an exchange.
- A license or white-label model lets a business buy rights, seats, deployment or control.
- Advertising or affiliate revenue means a third party pays for qualified attention.
Test the assumptions behind the model
- Who receives the value and who controls the budget?
- What observable event proves that value happened?
- Does serving the next customer cost almost zero or materially more?
- Would a customer prefer predictable spend or a price proportional to use?
Use metrics to expose different failure modes
| Metric | What it tests |
|---|---|
| Paid conversion by model | whether customers accept the proposed way of charging |
| Gross margin | whether revenue supports the variable cost of delivering the product |
| Revenue concentration | whether the model depends heavily on a small part of the customer base |
| Time to first value | how quickly the customer reaches the first observable value event |
Keep the test interpretable
Avoid copying a competitor whose cost structure is different, charging the user when another party receives the economic value, or launching several models before one has shown demand. Each choice makes it harder to connect customer behavior to the model being tested.
Set the success threshold before the experiment, then run the choice long enough to observe customer behavior. Keep what improves activation, retained revenue or contribution margin; stop what produces attention without those outcomes.