Industry survey data on household subscription spending, with the gap between recalled and actual figures explained through Soman's payment-friction research.
If you've ever sat down to total your monthly subscriptions, you've probably had the same experience most people have: you write down what you remember, you arrive at a number, you double-check against your bank statement, and the bank statement is larger. Often by a lot.
The gap between what people report paying for subscriptions and what they actually pay is one of the more interesting findings in consumer market research, and it's worth being honest about the source of that finding before we get into the numbers. The data on household subscription spending in North America comes almost entirely from industry surveys — Deloitte's annual Digital Media Trends, C+R Research's recurring subscription studies, Chase Bank's payment-data analyses, Antenna's streaming-industry reports. These are credible firms doing competent work, but they are not peer-reviewed academic research. The figures they produce should be read as directional rather than precise. The size of the gap they document, on the other hand, is robust across methodologies.
Across multiple years of these surveys, a consistent pattern appears. When American and Canadian households are asked to estimate their monthly subscription spending unprompted, the median answer is somewhere between $80 and $140 a month. When the same households are asked to itemize every recurring charge from their bank statements, the actual figure typically lands between $220 and $290. The gap — the part the household forgot about, or never noticed, or assumed had been cancelled — runs $80 to $150 a month for the typical respondent. Annualized, that's between a thousand and two thousand dollars of spending the household had not consciously budgeted for.
What's interesting about this gap is not its size, which is anecdotally familiar to anyone who has actually audited their own statements. What's interesting is that it exists at all. These are people, mostly, who are budget-aware. They use a card whose statement they receive monthly. They are, by definition, paying for these services every month. And yet the number they hold in their head is consistently smaller than the number on the statement.
To understand why, it helps to bring in some research that isn't market data. A 2001 paper in the Journal of Consumer Research by Dilip Soman established what he called the rehearsal hypothesis: that the cognitive act of explicitly engaging with a payment amount — counting bills, writing a check, signing a slip — is what determines whether the payment is later remembered. Paying by credit card weakens rehearsal substantially compared to paying by cash. Paying by automatic recurring billing — the default for nearly all subscriptions — removes rehearsal entirely.
Past payments strongly reduce purchase intention when the payment mechanism requires the consumer to write down the amount paid (rehearsal) and when the consumer's wealth is depleted immediately rather than with a delay (immediacy). — Soman (2001), Journal of Consumer Research
The gap between recalled and actual subscription spending is, in Soman's terms, the predicted outcome. The mechanism that would normally produce accurate spending memory has been deliberately and structurally removed from the experience.
The comparison to groceries — sometimes raised in the industry surveys — is worth handling carefully. The claim "the average household spends more on subscriptions than groceries" is mostly accurate for childless households without major food expenses, and mostly wrong for families of four with school-age kids who go through $400 of food a week. The framing is useful for shock value but misleading as a precise statistic. What's reliably true across household sizes: subscription spending is now within the same order of magnitude as grocery spending, where two decades ago it was an order of magnitude smaller.
The deeper observation in the financial well-being literature is that ambient uncertainty about money — not the dollar amount itself — is the dominant predictor of subscription-related stress. A 2018 paper in the Journal of Consumer Research by Netemeyer and colleagues distinguished current money-management stress from expected future financial security and showed both contribute to overall well-being independent of income. Subscription creep acts on both pathways. It generates current stress through the small monthly uncertainty ("am I paying for things I don't use?") and future stress through the unbounded trajectory ("what will my recurring bills look like a year from now?").
The intervention, if you want one, is mechanical. Pull the last 90 days of statements for every card the household uses. List every recurring merchant. Sum by category. The exercise takes about an hour. The number you arrive at is almost certainly different from the number you would have estimated unprompted. The decision about what to keep or cancel can wait; the number itself is the first piece of useful information.
What the research is fairly direct about — and what the audit consistently demonstrates — is that the friction of looking at the actual figure is the friction the system was designed to remove. Restoring it is, by Soman's mechanism, the act that restores spending awareness. Whatever you decide to do with the number afterward, the number itself is the change.