“When we build, add, invent, we rarely subtract.”
Leidy Klotz, Subtract: The Untapped Science of Less (2021)
A $3 increase is below your perceptual threshold. Ten $3 increases over 24 months is $30/month. $30/month is $360/year. $360/year is a return flight to London, or three months of groceries for a household of two, or a new laptop fund. The mechanism is deliberate. Pricing teams at Verizon, Comcast, Netflix, and Microsoft have studied the psychology of recurring charges with the same rigour that pharmaceutical companies study pain thresholds. They know the number that triggers cancellation. They stay just below it.
The term bill anaesthesia describes the condition in which recurring charges stop feeling real. Autopay is the mechanism: when money leaves your account automatically, without your conscious participation, the psychological cost of each transaction approaches zero. You feel your $150 dinner reservation. You do not feel your $87 streaming-and-subscription bill because it never required a decision to spend. It just happened, as it will next month and the month after.
Five behavioural levers that creep exploits
1. Default bias (Thaler and Sunstein)
In Nudge (2008), Richard Thaler and Cass Sunstein documented that people systematically choose the default option when a decision is required. Auto-renewal is the default. Cancellation is the opt-out. In an experiment on organ donation, moving from opt-in to opt-out systems increased consent rates from 4-27% to 85-99%. The same principle governs subscription renewals: the default is to pay again, and most people, most of the time, take the default. Signing up is an active choice; staying is not. The carrier or streaming service sets the default and collects the benefit.
2. Weber-Fechner threshold
The Weber-Fechner law in psychophysics states that the just noticeable difference between two stimuli is proportional to the baseline intensity. On a $120 cable bill, a $3 increase is 2.5%, well below the 10-15% threshold at which most people consciously notice a change. The same $3 increase on a $20 grocery item would be obvious. Pricing teams structure increases as a percentage of the bill, not as a flat dollar amount, because small percentages of large bills stay below the threshold of conscious notice for longer.
3. Loss aversion (Kahneman)
Daniel Kahneman's prospect theory (Thinking, Fast and Slow, 2011) establishes that the emotional impact of losing something is roughly twice the emotional impact of gaining the equivalent. When you contemplate cancelling a subscription, you are experiencing the anticipated loss of the service at a time when that service feels concrete and real, weighed against the abstract future saving of $15/month. The loss (giving up Netflix tonight) feels more visceral than the gain ($180/year you don't have yet). This asymmetry consistently biases decisions toward staying subscribed.
4. Status-quo bias
Staying subscribed requires no action. Cancelling requires a decision, a search for the cancellation path, often a phone call, sometimes a 20-minute hold, occasionally a retention offer that resets the decision. Every additional step in the cancellation process is an additional opportunity for status-quo bias to win. The FTC's Click-to-Cancel rule (effective November 2024) requires that cancellation must be as easy as sign-up. In practice, many carriers and streaming services interpret this minimally.
5. Temporal discounting
Behavioural economists have documented that humans systematically undervalue future costs relative to present costs. A $3/month creep that compounds over 24 months into $360/year in extra charges is emotionally invisible today because the harm is distributed across 24 future billing cycles. You do not experience the $360 loss as a lump sum in the way you would experience a single $360 charge. The same mechanism makes it hard to motivate action on climate, retirement savings, or preventive healthcare. The harm is real, accumulating, and emotionally unregistered.
The asymmetry
Signing up for a streaming service takes 30 seconds: enter your email, choose a plan, enter a payment method, done. Cancelling the same service requires navigating to Account, scrolling to find Manage Subscription or Membership, confirming through a retention quiz ("Why are you leaving?" "Would you like to pause instead?"), declining a promotional offer, and confirming cancellation. Netflix historically required account settings, membership and billing, and then a Cancel Membership buried under additional confirmation steps. Gym cancellations in many US states require written notice 30 days in advance. Planet Fitness requires a certified letter.
The onboarding / cancellation asymmetry is the subject of the signupdrop.com publication in this cluster. The relevant point here is that the asymmetry amplifies bill anaesthesia: the very act of attempting to cancel triggers a sequence of friction events that reinforce the status-quo bias and loss aversion documented above.
How to break the anaesthesia
The annual statement dump is the most effective intervention. Once per year, print every bank and credit card statement from the previous 12 months and read each line. The goal is to make each charge visible and to force a deliberate decision to continue or cancel. The physical act of printing and reading creates a psychological distance from the autopay default that reviewing a digital statement rarely achieves.
The renewal-calendar method creates a recurring audit at the moment of maximum leverage. Every time you subscribe to a service, immediately set a calendar reminder for 14 days before the next annual renewal date with a note: "Review: do I still use [service]? Is the price the same? Is there a cheaper tier?" 14 days gives you enough time to research alternatives, make a retention call, or cancel before the charge posts.
The bank-feed tool method is the lowest-friction option for people who will not do the manual audit. Rocket Money, Copilot Money, Monarch Money, and similar tools connect to your bank account and surface all recurring charges in one view, with the ability to cancel directly from the dashboard. The cost of Rocket Money Premium ($7-14/mo) is almost always less than the first month of savings it identifies.
The same mechanism at organisational scale
Everything described above applies at 10,000x the scale in SaaS procurement and cloud cost. The Zylo 2026 SaaS Management Index reports that 53% of enterprise SaaS spend is shelfware: tools that are paid for but not actively used. The reason is not negligence. It is the same five mechanisms. Default bias (the contract auto-renews). Weber-Fechner (a $2,000/month tool is a rounding error against a $1M SaaS budget). Loss aversion (cancelling might upset the team that used it once). Status-quo bias (it is on the renewal list and nobody questions it). Temporal discounting (the waste accumulates invisibly quarter by quarter).
The remedies are the same at scale: a systematic audit cadence, a centralised view of all recurring charges, and a named owner for each renewal decision. See SaaS spend audit and cloud cost optimisation for the enterprise versions of these interventions.