What is sunk cost fallacy and how does it work
The foundational research on why people continue commitments despite diminishing returns comes from psychologists Hal Arkes and Catherine Blumer. Their 1985 paper "The Psychology of Sunk Cost" in Organizational Behavior and Human Decision Processes established the empirical basis for what they termed the sunk cost effect: "a greater tendency to continue an endeavor once an investment in money, effort, or time has been made."
Their experiments demonstrated this effect with striking clarity. In one study, participants imagined purchasing tickets for two ski trips—a $100 trip to Michigan and a $50 trip to Wisconsin—only to discover both fell on the same weekend with no refunds possible. Despite being told the Wisconsin trip would be more enjoyable, approximately 50% of participants chose the more expensive Michigan trip. The rational choice was clear, but the psychological pull of the larger sunk cost proved equally compelling.
A field experiment at a theater reinforced these findings. Patrons randomly received either full-price tickets ($15), modest discounts ($2 off), or substantial discounts ($7 off). Over the season, those who paid full price attended significantly more performances than those who received discounts—even though the entertainment value was identical. The money was spent regardless; only the psychological weight of that expenditure differed.
Recent research has extended these findings specifically to digital subscriptions. A 2025 study in the Journal of the Association for Information Systems found that subscription fee payments create sunk costs that directly influence consumption behavior, with box office revenues increasing 12-35% following subscription price adjustments. Similarly, 2022 research from Wharton faculty determined that only one-third of increased purchasing from subscription programs stems from economic benefits—the remaining two-thirds is attributable to non-economic factors including sunk cost fallacy.
Loss aversion: why losing feels twice as painful
The theoretical framework explaining sunk cost psychology comes largely from Kahneman and Tversky's prospect theory, published in Econometrica in 1979. Their central insight was deceptively simple: losses and gains of equal magnitude do not feel equal. Losing $100 hurts more than gaining $100 feels good.
How much more? Tversky and Kahneman's 1992 follow-up research estimated a loss aversion coefficient of approximately 2.25—meaning losses are experienced roughly twice as intensely as equivalent gains. This asymmetry has profound implications for subscription cancellation. When contemplating ending a service, the brain processes this as an impending loss: loss of access, loss of content, loss of convenience. The potential financial savings—a gain—registers with roughly half the psychological intensity.
This explains why subscription services emphasize what users will lose upon cancellation. Messages highlighting "You'll lose access to 10,000 movies" or "Your curated playlists will be deleted" trigger loss aversion responses that dwarf the abstract appeal of saving $15 monthly. The framing isn't accidental; it's psychologically optimized.
The phenomenon interacts with another documented bias: the endowment effect, first named by economist Richard Thaler in 1980. Classic experiments by Kahneman, Knetsch, and Thaler demonstrated that people demand significantly more to part with objects they own than they would pay to acquire the same objects. In the famous Cornell mug studies, sellers required roughly $7 to part with a mug, while buyers would pay only about $3 for the identical item—a willingness-to-accept/willingness-to-pay ratio of approximately 2:1.
Applied to subscriptions, this means users develop a sense of ownership over their Netflix queue, their Spotify playlists, their gym membership identity. The personalization algorithms that learn preferences deepen this ownership attachment. Canceling feels less like ending a service contract and more like giving away something that belongs to you.
Status quo bias and mental accounting compound the effect
Beyond sunk costs and loss aversion, subscription retention benefits from status quo bias—the documented preference for maintaining current states even when alternatives might be objectively superior. Samuelson and Zeckhauser's 1988 research in the Journal of Risk and Uncertainty quantified this effect: an option gains approximately 17 percentage points simply by being designated as the current choice.
Their field studies revealed the bias operating in consequential real-world decisions. Among Harvard employees with health plan options, only 3% switched plans annually, with the incumbent plan chosen by 2-4 times more existing employees than new enrollees across age groups. In TIAA/CREF retirement accounts, fewer than 28% of participants had ever changed their allocation, and under 2.5% adjusted annually despite substantial market fluctuations. People stick with what they have, even when they chose it years ago under different circumstances.
Auto-renewal transforms this inertia into subscription retention. The status quo becomes continued payment; cancellation requires active effort against the default. Research supports the magnitude of this effect: studies found that approximately 8% of customers cancel when required to actively renew, versus only 2% during auto-renewal months—a fourfold difference driven purely by the direction of the default.
Thaler's research on mental accounting adds another dimension. People don't treat money as fungible; instead, they mentally categorize funds into separate accounts—housing, entertainment, subscriptions. Once a "streaming budget" mental account exists, a $15 monthly charge seems small against a $100 allocation, even if the objective question is whether that $15 delivers proportionate value.
The gym membership paradox illustrates subscription psychology
Perhaps no subscription better demonstrates these psychological dynamics than gym memberships, where the gap between payment and usage reaches almost absurd proportions. Research indicates 67% of gym memberships go completely unused—members pay but never attend. The average gym membership lasts 4.7 years despite this low utilization, and Americans collectively waste approximately $1.3 billion annually on gym memberships they don't use.
The temporal patterns are equally revealing. IHRSA (International Health, Racquet & Sportsclub Association) research shows 12-12.5% of annual gym sign-ups occur in January, versus an 8.3% monthly average—the New Year's resolution surge. Yet 80% of January joiners quit within five months. The industry has even coined "Fall Off the Wagon Day" for the second Saturday in February, when attendance returns to baseline levels.
The business models of major gym chains explicitly depend on non-attendance. NPR reporting revealed that Planet Fitness locations average 6,500 members but can only accommodate 300 people simultaneously—an overselling ratio of more than 20:1. The economic model assumes most members will continue paying while staying home, and behavioral economics explains why this assumption consistently proves correct.
A RunRepeat survey found that 23% of members who eventually quit cite not using their membership enough as the reason—implying that 77% continue paying even when usage has dropped below their own satisfaction threshold. The sunk cost of previous payments, combined with optimistic beliefs about future usage and the friction of cancellation, keeps payments flowing long after workouts have stopped.
Streaming services and the phenomenon of subscription stacking
The streaming landscape presents a different subscription psychology pattern: stacking multiple services while cycling usage between them. Current data shows the average U.S. household subscribes to 4.1 paid video streaming services, with 53% maintaining at least four subscriptions simultaneously. Combined monthly streaming costs average $61-69 across these services.
The streaming industry exhibits notably lower churn than other subscription categories, with Netflix maintaining the lowest rate at 1.8-2.4% monthly gross churn. Industry-wide Premium SVOD (subscription video on demand) churn averages around 5-7%. But the more interesting pattern is re-subscription behavior: 50% of Netflix subscribers who canceled in 2023 returned within six months, and 61% rejoined within a year. The industry-wide median for subscribers returning to a canceled service reached 34.2% in 2024, up from 29.8% in 2022.
This "subscription cycling" behavior—canceling and re-subscribing as content interests shift—represents a more sophisticated consumer response to subscription fatigue, which surveys indicate affects 42% of subscribers who feel they have too many streaming services. Simon-Kucher research found nearly 50% of subscribers plan to cancel at least one service within the next year, though many will likely return.
The underestimation problem persists in streaming as in other subscriptions. C+R Research documented that consumers believe they spend approximately $86 monthly on subscriptions but actually spend $219—a 2.5x gap.
Free trials and cancellation friction exploit psychological vulnerabilities
The subscription economy's growth has been accompanied by increasingly sophisticated exploitation of these psychological mechanisms, documented by academic researchers and regulatory bodies alike.
Free trials leverage the endowment effect by establishing psychological ownership before any payment occurs. During trial periods, users customize preferences, build playlists, and accumulate viewing history—investments that create sunk costs of time and effort even without monetary payment. Survey data reveals 70% of people who sign up for free trials with the intention of canceling have forgotten to do so at least once. Opt-out trials requiring credit card information convert at approximately 48.8-60%, compared to roughly 25% for opt-in trials requiring no payment information.
Cancellation friction—deliberately complicated termination processes—has attracted substantial research attention. Princeton researchers conducting the largest study of dark patterns crawled 11,000 shopping websites and identified 1,818 dark pattern instances across 1,254 websites, with 183 engaging in explicitly deceptive practices. They catalogued patterns including "sneaking" (hidden charges or auto-enrollments), "misdirection" (attention-directing design), and "obstruction" (difficult-to-execute actions like cancellation).
The FTC now receives nearly 70 consumer complaints daily about negative option and subscription practices, up from 42 daily in 2021. Their investigation of major subscription sites found 81% used "sneaking" tactics, 70% didn't provide cancellation information, and 67% failed to disclose cancellation deadlines. The agency's 2023 lawsuit against Amazon alleged the Prime cancellation process was internally nicknamed "the Iliad"—requiring 4 pages, 6 clicks, and navigation through 15 options to complete.
Consumer surveys reflect these difficulties: 76% of people believe subscription services make cancellation deliberately difficult, and 60.4% report having avoided signing up for services specifically because they anticipated cancellation problems.
How expense tracking counteracts psychological biases
Interestingly, behavioral economics research also points to ways of mitigating these psychological effects. One of the most effective approaches has proven to be simple expense tracking—systematic recording of where money goes. When abstract monthly payments become concrete numbers in categories, the psychological distance between decisions and their consequences shrinks.
Visualizing subscription spending in the context of an overall budget helps counteract the mental accounting effect. Instead of $15 for streaming seeming small within a $100 "entertainment budget," tracking shows how four such subscriptions combine into $60—nearly a third of grocery spending for some households. This shift in perspective doesn't dictate decisions, but it changes the informational basis for making them.
MyFin is designed precisely for this type of awareness without providing financial advice. The app tracks regular transactions and automatically identifies recurring payments—those same subscriptions that are easy to forget. Categorization shows what portion of a budget goes to subscriptions, while time-based visualization demonstrates how these costs accumulate over months and years. For people who value control over their data, MyFin operates locally without transmitting financial information to the cloud—an important characteristic for users who prioritize privacy.
Research shows that even passive observation of spending changes behavior—a phenomenon known as the "observation effect" in psychology. When people know their spending is being recorded and will be reviewed, they naturally become more selective. In the context of subscriptions, this means that regularly reviewing a "subscriptions" category in a tracking app creates reassessment points that auto-renewal is designed to avoid.
Conclusion: understanding the mechanisms doesn't eliminate their influence
The psychology of subscription retention operates through multiple, reinforcing mechanisms: sunk cost fallacy converts past payments into pressure to continue; loss aversion makes cancellation feel twice as painful as the equivalent savings feel beneficial; status quo bias favors maintaining existing arrangements; the endowment effect creates attachment to personalized service features; and mental accounting compartmentalizes subscription spending in ways that reduce scrutiny. These aren't bugs in human cognition—they're features that evolved for valid reasons, now operating in an economic environment that systematically exploits them.
The subscription economy has grown 435% over the past nine years, substantially outpacing broader market growth, and is projected to reach $1.5 trillion by 2025. This growth reflects genuine value creation—convenient access models that consumers prefer to ownership. But it also reflects an industry that has become increasingly sophisticated at leveraging behavioral economics research to optimize retention. The FTC's 2024 "click-to-cancel" rule represents regulatory recognition that market forces alone haven't corrected the resulting imbalances.
Understanding the psychological mechanisms doesn't eliminate their influence, but it does explain why that unused subscription feels so hard to terminate—and why the difficulty isn't really about remembering the cancellation deadline. Expense tracking tools offer a practical way to create counterweights to these psychological forces—not by imposing decisions, but by providing clarity about how money is actually being spent. In a world that's getting increasingly better at exploiting cognitive features to its advantage, awareness becomes the most valuable form of protection.
