Research shows that forgotten renewals and obstructive interfaces can increase subscription earnings—while a new US lawsuit puts the business model under scrutiny.
Signing up for software can take seconds. Leaving may mean searching through settings, contacting support, answering retention questions and discovering a fee that was difficult to spot at checkout. The imbalance is familiar to anyone who has kept paying for an app they stopped using, but it also reveals a deeper design problem: some recurring revenue depends on customers failing to leave.
On October 29, the US Federal Trade Commission and state authorities sued restaurant software provider Toast, alleging hidden charges and difficult cancellation practices. The case moves a largely invisible product-design decision into a public legal dispute: how much effort should a company be allowed to demand before ending a subscription?
When Software Becomes Difficult to Leave
Toast sells restaurants cloud-based software for payments, orders, payroll and other operations, often through automatically renewing contracts. The FTC alleges that cancelling requires 30 days’ written notice, followed in some cases by additional customer-service contacts and sales pitches. The agency also describes an undisclosed early termination charge of $150 for every remaining contract month; one customer reported a $5,318.87 cancellation fee.
Financing can complicate the exit. According to the complaint, cancelling Toast services may trigger immediate repayment of outstanding Toast Capital loans. These are allegations, not court findings.
Toast disputes the claims. In its response, the company argues that the federal law cited by regulators was designed for online subscriptions, not negotiated business-to-business agreements, and says nearly 200,000 business locations use its platform. The litigation is pending.
The case matters beyond restaurants because the same commercial logic can appear in consumer software: joining is optimized for conversion, while leaving may be managed as a retention opportunity.
The Economics of People Who Forget
An influential 2025 American Economic Review study by economists Liran Einav, Ben Klopack and Neale Mahoney investigated why customers keep paying for services they no longer value. Using transaction data from a large payment-card network, the researchers found cancellation increased when replacement cards forced subscribers to make an active renewal decision.
Their models estimated that consumer inertia—through inattention or switching costs—roughly doubled seller revenue on average, holding the initial subscriber base fixed.
That is a model-based finding from the subscriptions studied, not proof that all software providers intentionally obstruct cancellation. Yet it exposes the incentive: if an inactive user continues paying, the company receives revenue without having to persuade that person the product remains worthwhile.
For a software business reporting monthly recurring revenue, this raises an uncomfortable measurement question. How much retention reflects continuing satisfaction, and how much is simply the absence of a cancellation?
Dark Patterns Turn Interface Design Into Pressure
The Organisation for Economic Co-operation and Development describes dark commercial patterns as interface practices that can steer or manipulate people toward decisions against their interests. Examples include preselected options, hidden information and difficult-to-find cancellation controls.
The scale is significant. An international 2024 review coordinated by the International Consumer Protection and Enforcement Network examined 642 subscription websites and apps across multiple countries. Nearly 76% showed at least one potential dark pattern, while nearly 67% displayed multiple. The assessment did not determine whether those practices were illegal.
An open-access study published in Marketing Letters by Julian Runge, Daniel Wentzel, Ji Young Huh and Allison Chaney offers a closer look at friction. Across the services they examined, account creation averaged 18 clicks and 1.7 minutes; deletion averaged 27 clicks and four minutes.
Deleting an account is not identical to cancelling a paid subscription, and the study was observational. Still, it demonstrates how design can make departure more demanding than entry.
The Product Metrics That Can Reward Friction
Digital subscriptions give legitimate benefits: predictable pricing, automatic updates and continuous access. Recurring revenue can also help developers maintain security, support and product improvements.
The problem arises when retention becomes the objective regardless of customer intent. A team rewarded primarily for reducing churn might be encouraged to add confirmation screens, hide an exit option or route cancellation through a sales conversation.
That can make dashboards look healthier without proving the service delivers more value. Engineering teams also inherit the complexity: billing state, notice periods, account permissions and refund handling must agree, or users may remain charged after believing they have left.
A more meaningful measurement approach would distinguish voluntary renewals from inactive subscriptions, track failed cancellation attempts and test whether people can exit without unnecessary intervention.
The New Test of Software Trust
The OECD warns that dark patterns can impose financial costs, undermine privacy and weaken competition. For founders, the alternative is not abandoning subscriptions; it is treating a transparent exit as part of product quality.
Clear renewal disclosures, visible cancellation controls, immediate confirmation and transparent fee calculations could reduce accidental payments. They might also expose revenue previously hidden inside friction.
That is the trade-off the Toast case and economic research bring into focus. Subscription software succeeds legitimately when customers repeatedly choose its value. When keeping a customer requires making departure difficult, a product’s recurring revenue may reveal more about its exit design than its usefulness.
