The difference is not ethics. It is whether the customer's success state is exit or continued use.
Most subscription businesses in PE portfolios are evaluated as if the model is a given and the operating question is execution: reduce churn, raise NRR, expand seats.
For some portcos, that is the right question. For others, it is the wrong question, asked with increasing intensity.
The diagnostic that sits upstream of churn, NRR, and expansion is older. Is the customer's underlying problem recurring or terminal?
Recurring problems
A recurring problem generates recurring desire. The customer wants something on an ongoing basis: fresh content, current data, continued access, regular insight. The subscription model lets the provider deliver against that desire continuously. The provider wins by working. The customer wins by being served. Retention becomes a property of the product, not of the retention team.
OnlyFans, Bloomberg Terminal, Spotify, a good industry research subscription. Different categories, same alignment. The customer's success state is continued use.
Terminal problems
A terminal problem generates the opposite dynamic. The customer wants to resolve something: get fit, get out of debt, get hired, find a partner, learn a language, fix a broken process. Their success state is exit.
When a subscription is wrapped around a terminal problem, the provider's revenue depends on the customer's failure to resolve it. Over time, the economics force a product that delays resolution rather than enables it.
This is not a hypothesis. DellaVigna and Malmendier (2006), studying U.S. gym contracts in the American Economic Review, showed how providers profit-maximise against consumers' overconfidence about future self-control: pricing flat-rate access far above per-visit cost and making cancellation deliberately frictional. In August 2025, the FTC settled with Match Group for $14M over fake love-interest ads and obstructed cancellation flows on Match.com and Tinder. The pattern repeats wherever the underlying problem is terminal: engagement metrics rise, resolution metrics fall, and eventually the cohort notices.
How this shows up in B2B
The same dynamic runs in B2B, less visibly. A SaaS subscription whose value depends on the customer never quite finishing the implementation. A platform whose dashboard never quite produces the answer. A managed-service contract whose scope expands each renewal because the original problem was never closed.
The diagnostic question for an OP
At the model layer, the question is simple. Does the customer's success state coincide with continued subscription, or with exit?
If continued subscription: the model compounds. NRR work, expansion motion, and CS investment all return.
If exit: the model is structurally misaligned. The operating teams are running retention strategies against a customer who, at some level, is trying to leave. The dashboard reads as an execution problem. It is a model problem.
The portcos that compound on subscription revenue are the ones where staying is the win-state for both sides. The ones that struggle quietly are the ones where the customer's success and the company's revenue point in opposite directions.
No amount of operating intervention will close that gap, because the gap is the model.
References
- DellaVigna, S., & Malmendier, U. (2006). Paying not to go to the gym. American Economic Review, 96(3), 694–719.
- Federal Trade Commission. (2025, August 12). Match Group agrees to pay $14 million, permanently stop deceptive advertising, cancellation, and billing practices to resolve FTC charges [Press release].
- Visnjic, I., Jovanovic, M., Neely, A., & Engwall, M. (2017). What brings the value to outcome-based contract providers? Value drivers in outcome business models. International Journal of Production Economics, 192, 169–181.