The Subscription Trap: How Recurring Billing Became a Default Setting

Somewhere in the last two decades, the default structure of consumer commerce inverted. Buying something used to mean an affirmative act repeated each time you wanted it. Increasingly it means one affirmative act followed by indefinite consent, where the charge continues until you successfully intervene — and where the difficulty of intervening is a variable the seller controls. That structure has a name in regulatory language, negative option billing, and a straightforward economic logic: revenue from customers who no longer want the product is the highest-margin revenue available, because it costs nothing to serve someone who has stopped using what they're paying for. This report examines the mechanics, the current legal position after a federal rule was struck down, and what actually works.

The core thesis

Subscription commerce is usually analyzed as a business model innovation — predictable revenue, better customer relationships, lower acquisition costs amortized over a longer life. All of that is true and much of it is genuinely good for consumers, who get lower entry prices and continuous product improvement.

Our thesis concerns the part that isn't discussed: the model's most profitable customer is one who has stopped valuing the product but hasn't stopped paying, and that fact shapes design decisions throughout the funnel. Not because subscription companies are unusually cynical, but because the metric they optimize — retention — is measured identically whether it results from satisfaction or from inertia. A customer who stays because the product is good and a customer who stays because cancellation requires a phone call during business hours both appear in the same cell of the same dashboard.

That measurement equivalence is the mechanism. Once retention is the objective and friction is a lever that moves it, the lever gets pulled — incrementally, defensibly, and usually without anyone framing it as harming customers. The result is the pattern regulators describe as an asymmetry between enrollment and cancellation: one click to start, a sequence of obstacles to stop.

The second half of the thesis is about who bears it. Subscription drag falls hardest on households with the least slack — the ones our buffer analysis describes, where an unnoticed recurring charge is the difference between clearing and overdrawing. It's a small version of the same regressive structure our poverty premium report documents: a cost that attaches to inattention, and inattention is cheapest to avoid for people with time and financial cushion.

Retention is measured identically whether it comes from satisfaction or from inertia — which is why friction becomes a lever, and why it gets pulled.

The economics of forgetting

Three properties make unused subscriptions structurally profitable.

Marginal cost approaches zero for a non-user. A customer who stopped opening the app consumes no support, no bandwidth, no fulfillment. Their payment is nearly pure margin, which means the revenue from forgotten subscriptions is worth considerably more per dollar than revenue from active ones.

The amounts are individually small and collectively large. A charge sized below the threshold at which a household scrutinizes its statement can persist for years. Across a customer base, those charges are material; for the individual, each one is beneath the effort of investigating.

Payment is invisible by design. The card on file, the automatic renewal, and the account updater services that keep credentials current through card reissuance all mean the transaction requires no attention — and modern tokenization means even replacing a card frequently doesn't stop it, since network updater services carry the subscription forward. The infrastructure built to reduce the involuntary churn our failed payments analysis examines also makes deliberate exit harder.

Put together, these produce a business incentive that doesn't require bad intent to operate: every friction that delays cancellation converts directly into high-margin revenue, and the delay need only be measured in billing cycles to pay for itself.

Friction as a designed asset

The catalogue of cancellation obstacles is consistent enough across industries to be described as a pattern language:

• Channel asymmetry. Sign up online in seconds; cancel by telephone, during limited hours, with a wait.
• Retention gauntlets. A sequence of offers, surveys, and confirmations, each requiring a separate action, where the cancel path is visually subordinate to the "stay" path at every step.
• Buried controls. Cancellation located several menus deep, under a label that doesn't say cancellation.
• Confirmation ambiguity. A flow that ends without clearly stating that the subscription is cancelled, leaving the consumer unsure whether it worked — and frequently it hasn't.
• Notice requirements demanding cancellation a set period before renewal, so a customer who acts on the renewal date has already renewed.
• Pause offers substituted for cancellation, which resume automatically.
• Silent renewal at a higher price, where a promotional rate converts without meaningful notice.

Two things are worth saying precisely about this list. First, each item is individually defensible — a retention offer is a legitimate business practice, phone cancellation exists for identity verification, and notice periods have operational rationales. It's the asymmetry that regulators have focused on: the gap between how easy it is to start and how hard it is to stop. Second, the practices are not confined to disreputable operators. Some of the most familiar consumer brands have faced enforcement over cancellation flows, which is what makes this a structural analysis rather than a bad-actor story.

Why free trials work

The free trial is the negative option structure in its purest form, and its effectiveness rests on well-understood behavioral mechanics.

The decision is displaced in time. At signup, the cost is zero and the future charge is abstract. At conversion, the customer has already integrated the product, and cancelling now feels like a loss rather than a decision not to buy.

The endowment effect operates. Having used the service, the customer values it more than they did before trying it — which is partly genuine information and partly a documented cognitive bias the structure exploits.

The friction arrives at the worst moment. Trial cancellation frequently requires more steps than trial signup, and the customer is now making that effort to avoid a charge rather than to obtain a benefit — a weaker motivation.

And notice is minimal. A trial converting without a clear pre-charge reminder relies entirely on the customer's own memory and calendar.

None of this makes trials illegitimate. A trial is genuinely useful for products whose value can't be assessed from a description, and plenty of customers convert happily. The distinction that matters — and the one both regulators and honest operators draw — is between a trial that informs a decision and a trial engineered so the decision never gets made. The tell is whether the company sends a reminder before charging. One that does is selling the product; one that doesn't is selling the forgetting.

The rule that was vacated

The federal regulatory story is recent, specific, and frequently misreported, so it's worth stating precisely.

In October 2024 the FTC finalized amendments to its Negative Option Rule, widely known as the click-to-cancel rule. It required sellers to disclose material terms clearly, obtain express informed consent before charging, and — the provision that gave the rule its name — provide a cancellation mechanism at least as simple as the enrollment mechanism. If you could subscribe with one tap in an app, you had to be able to cancel with one tap in the app.

The rule was set to take effect in July 2025. On July 8, 2025, the Eighth Circuit vacated it in full in Custom Communications, Inc. v. FTC. The grounds were procedural rather than substantive: the court held that the Commission failed to conduct a preliminary regulatory analysis required under the FTC Act once it became apparent the rule would impose more than $100 million in annual compliance costs, and that the procedural failure was prejudicial.

What followed matters as much as the vacatur. The FTC restored the Negative Option Rule to its pre-2024 version in early 2026, and in the same period initiated a new rulemaking, submitting an advance notice of proposed rulemaking seeking comment on how to address recurring charges consumers didn't intend and cancellation obstacles. Commission leadership has publicly reaffirmed a focus on deceptive negative option subscriptions. Given that the prior rule took roughly three years from proposal to final, and given the procedural care the Commission will now take, a replacement is a multi-year prospect rather than an imminent one.

The episode is a useful case study in something this desk has noted elsewhere: a rule can be defeated on procedure without the underlying conduct becoming lawful. The same pattern appeared in the consumer data rulemaking our broker analysis examines — the perimeter moved, the enforcement authority didn't.

What still applies

The most consequential misreading of the vacatur is that subscription cancellation is now unregulated. It isn't.

Authority	What it covers
ROSCA (Restore Online Shoppers' Confidence Act)	Online negative option offers: clear disclosure of material terms, express informed consent before charging, and a simple mechanism to stop recurring charges. The FTC has interpreted this to require cancellation at least as easy as enrollment, and has brought cases under it across administrations.
FTC Act Section 5	Unfair or deceptive acts or practices generally — the authority the Commission has continued to use against cancellation practices, including where the cancellation path is materially harder than the enrollment path.
1973 Negative Option Rule	Still in force, governing prenotification plans for goods shipped by mail or telephone order.
State automatic renewal laws	Enforceable and in several states comparable to or stricter than the vacated federal rule, with meaningful penalties.
State UDAP authority	State attorneys general applying unfair and deceptive practices statutes, an active enforcement channel.

There's an additional practical point that businesses should register: the vacated rule functions as a benchmark even though it isn't law. Plaintiffs' counsel and regulators can and do reference the symmetry standard as the reasonable practice, which means a company whose cancellation flow would have failed the 2024 rule is not in a comfortable position simply because the rule was struck down.

The state layer

With federal rulemaking restarted rather than resolved, state automatic renewal laws are the operative regime for most subscription businesses — and the pattern rhymes with what our licensing analysis describes across financial regulation: fragmentation as the default, with the strictest state effectively setting national practice for anyone operating everywhere.

California, New York, and Massachusetts are among the states with specific negative option or automatic renewal provisions, and others have enacted or considered similar measures. Common requirements across these regimes include clear and conspicuous disclosure of renewal terms before enrollment, affirmative consent, advance notice before renewal in defined circumstances, and cancellation mechanisms that don't require substantially more effort than signup. Several carry statutory damages that make private litigation viable.

The practical consequence for a business operating nationally is that complying with the strictest applicable state is usually cheaper than maintaining state-specific flows, which means the vacated federal rule's substance survives in large part through state law and voluntary adoption — a common outcome when federal action stalls.

The household ledger

Stepping back from the regulatory layer, the household-level phenomenon is simply that recurring charges accumulate faster than they're reviewed.

The mechanism is straightforward. Each individual subscription is small enough to approve without much thought. There is no natural review point — nothing forces an annual reckoning the way a renewal notice on an insurance policy does. Charges hit a card statement among dozens of other lines. And the cognitive cost of auditing them exceeds the cost of any single one, which means the rational individual decision is to not bother, repeated until the aggregate is substantial.

Two consequences worth noting. For households near the margin, subscription drag interacts directly with overdraft exposure — an unexpected recurring charge on a low balance triggers the fee cascade our overdraft analysis documents, meaning a small forgotten subscription can cost a multiple of its own value. And the audit is genuinely high-return: reviewing twelve months of statements once a year is among the highest hourly-return financial activities available to most households, precisely because nothing else forces the review.

How to actually cancel something

1. Run the audit properly. Twelve months of card and bank statements, not one — annual subscriptions hide from a monthly review. List every recurring charge with amount and frequency.
2. Cancel through the company first, and document it: date, method, screenshots of confirmation pages, names and reference numbers from calls.
3. Send written notice if the flow is obstructive — to the address in the terms — and keep proof of sending. This creates a record that matters later.
4. Watch the next billing cycle. A cancellation that doesn't stop the charge is common enough that verification should be part of the process.
5. Dispute continuing charges with your issuer as unauthorized recurring transactions, providing your cancellation documentation. Separately, instruct the issuer to block future charges from that merchant — a distinct step from disputing past ones, and the one people miss.
6. Don't rely on replacing the card. Account updater services frequently carry subscriptions to a reissued number, which is why an explicit stop instruction works better than a new card.
7. Complain formally to the FTC and your state attorney general. This creates the enforcement record that drives cases, and in practice it frequently produces a resolution on its own.
8. Prevent recurrence: a calendar reminder a few days before any trial converts, a dedicated card or account for subscriptions so the ledger is legible, and a standing annual review date.

Scenarios and what we're watching

Scenario	Shape of the world	Signposts
Base case — states lead	Federal rulemaking proceeds slowly; state automatic renewal laws and ROSCA enforcement set practice; large operators standardize on the strictest state	New state ARL enactments; FTC Section 5 and ROSCA case volume; rulemaking milestones
Federal replacement case	A procedurally durable rule restores the symmetry standard nationally, with clearer consent and notice requirements	NPRM publication; scope of proposed requirements; compliance analysis completion
Drift case	Absent a federal standard, cancellation friction expands in less-regulated categories while enforcement concentrates on the largest brands	Complaint volumes; state enforcement actions; disclosure practice across smaller operators

What we're watching: the pace and scope of the new federal rulemaking, particularly whether a symmetry requirement returns; state automatic renewal law adoption, which is where the operative standard currently lives; enforcement volume under ROSCA and Section 5, which indicates whether the vacatur changed conduct or only the citation; and platform-level intervention, since app stores and card networks can impose cancellation standards on merchants faster than any regulator. The negative option structure isn't going away — it's genuinely efficient for products people want continuously. The open question is whether the exit stays as easy as the entrance, and right now that's determined by which state you live in and which brand you subscribed to.

Frequently asked questions

---

This report is for general information only and does not constitute legal advice. The regulatory position described reflects publicly reported developments as of publication and is actively changing; verify current federal and state requirements before relying on any description here.