Subscription Services: Why We Keep Subscribing (Even When We Shouldn’t)

There’s a well-documented phenomenon in behavioural finance called the “dead subscriptions problem.” Research consistently finds that people significantly underestimate how many subscription services they’re paying for, and significantly overestimate how frequently they use them. A study by C+R Research found that the average American spends nearly $220 per month on subscriptions — roughly double their estimate when asked.

This isn’t surprising when you look at the mechanics. Subscriptions are engineered to be low-friction to start, automatic to continue, and — through a combination of inertia, inconvenience, and optimism about future usage — easy to keep even when the value has stopped being there.

That said, subscriptions are also genuinely one of the best consumer models available for many product categories. The question worth asking isn’t “are subscriptions bad?” but “which subscriptions are actually worth it, and for whom?”

Why the Subscription Model Took Over

From the provider’s perspective, the subscription model solves a fundamental problem with one-time purchases: revenue is lumpy, customer relationships end at the moment of sale, and growth requires constantly finding new buyers.

Subscriptions replace this with a predictable, recurring revenue stream. Investors love it — the metrics are predictable, churn can be measured and managed, and customer lifetime value is much easier to calculate. Adobe, Microsoft, Spotify, Netflix, and thousands of SaaS companies have built substantial valuations on the model’s economics.

For consumers, the advantages are genuine in many cases:

Continuous access to updated products: Software subscriptions mean you always have the latest version. This matters more as software has become more complex and security-critical — running a five-year-old version of something like a design tool or productivity suite is increasingly problematic, and the old model of paying for major version upgrades solved this imperfectly.

Lower upfront cost: The entry barrier to a subscription is much lower than an equivalent one-time purchase. Accessing Adobe Creative Suite on a subscription for £54/month is different from buying it outright at what was historically several thousand pounds. For individuals, freelancers, and small businesses, that accessibility genuinely matters.

Flexibility: Most subscriptions can be cancelled. If you need a service for three months and then don’t, you can pay for three months and stop. Ownership models don’t offer this.

Shared infrastructure: Services like cloud storage, streaming platforms, and developer tools pool infrastructure costs across millions of subscribers, making per-user costs lower than any individual could achieve independently.

The Mechanics of Subscription Stickiness

Understanding why subscriptions are so sticky is useful both for avoiding the ones that don’t serve you and for appreciating when stickiness is actually a feature.

Habit formation: Daily-use products that become part of a workflow are genuinely difficult to replace. Productivity tools, communication platforms, and code editors integrate deeply into how people work. Cancelling isn’t just a billing decision — it’s a workflow disruption.

Switching costs: Calendar integrations, stored preferences, connection history, proprietary file formats. Services deliberately create these switching costs to increase the friction of leaving. Sometimes this is incidental (your data is just stored in their format); sometimes it’s deliberate.

The cancellation UX: Most subscription services make cancelling deliberately non-trivial. They require navigating multiple screens, surviving a “please stay” flow, waiting through “are you sure?” dialogues, and sometimes — in what regulators have started to flag — making you call a phone number during business hours. The FCA and FTC have both increased scrutiny of cancellation friction, but it remains common.

Anchoring on sunk cost: “I’ve been paying for this for two years, I should start using it” is not a rational basis for continuing a subscription. But it’s a very common one.

Loss framing: Services that advertise “don’t miss out on [upcoming feature]” or “your membership benefits expire if you cancel” are framing cancellation as losing something, which activates loss aversion more effectively than framing continued membership as a gain.

Where Subscriptions Deliver Genuine Value

Cloud storage (iCloud, Google One, Dropbox): For people who use multiple devices, want automatic backup, and share large files, cloud storage subscriptions are often very good value. The per-GB cost is low, the reliability is excellent, and the convenience of not managing local storage on every device is real.

Creative and professional software (Adobe, Figma, Sketch): For people who use these tools professionally, subscription pricing is clearly justified — the tools are essential, they’re continuously updated with features that matter, and the alternative of buying perpetual licenses was expensive and created versioning problems. For occasional users, the same subscriptions are often poor value.

Streaming media: The calculus here has shifted as the number of services has multiplied. In 2019, Netflix, Spotify, and perhaps one other service could cover most media consumption at reasonable total cost. In 2024, content is fragmented across Netflix, Disney+, Apple TV+, Amazon Prime, Paramount+, and multiple sport streaming services. The aggregate cost of accessing everything across all of them often exceeds what cable packages used to cost — with less convenience, because you have to switch between apps. Rotating services rather than holding all of them simultaneously has become a legitimate strategy.

Security and utility software (password managers, VPNs, backup tools): Generally good value because the product genuinely requires ongoing infrastructure (servers, security updates, synchronisation). 1Password, Backblaze, and similar tools charge subscription prices that reflect actual costs and deliver continuous value.

Where Subscriptions Should Be Scrutinised More Carefully

Gym memberships: The most famous example of a subscription used far less than anticipated. Most people overestimate how frequently they’ll go. The pricing model depends on this — gyms would be unprofitable if every member showed up as often as they plan to when signing up.

News subscriptions: Genuine value for news sources you read daily. Poor value for publications you open twice a year when someone sends you a link. The subscription-based news model has legitimate merits for journalism funding; the consumer decision should be based on actual reading frequency.

SaaS tools you “might use”: The enterprise software landscape now includes hundreds of SaaS products that are genuinely useful for specific purposes. The problem is accumulation — teams sign up for tools for a specific project, the project ends, and the subscriptions run on because nobody tracks them. Tools like Cledara, Zluri, and Torii exist specifically to address this in organisations, which tells you how significant the problem is.

“Premium” upgrades for free products: Free products that push users toward paid tiers often use feature restrictions (limited storage, watermarks, usage caps) that create artificial pressure. Sometimes the upgrade is genuinely justified; sometimes it’s designed to make the free tier barely functional.

A Practical Audit Process

The only reliable way to know what you’re paying for is to look at your bank statements rather than trusting your memory. A quarterly subscription audit — filtering transactions by recurring billing — typically reveals at least one service most people had forgotten about.

For each service found:

1. When did you last use it?
2. Would you re-subscribe today if your subscription had lapsed?
3. Is there a cheaper tier that covers your actual usage?

The second question is the most honest filter. People continue subscriptions they wouldn’t actively choose to start, because inertia is the path of least resistance. Reversing the frame — “would I subscribe to this now?” — removes the sunk cost bias.

The services that survive that filter are worth keeping. The ones that don’t are either candidates for cancellation or for a downgrade to a lower tier that actually matches usage.

The Subscription Economy Is Not Going Away

The model works too well for providers to retreat from it, and enough subscriptions deliver genuine value that consumer rejection hasn’t materialised. What has happened — particularly in the streaming and SaaS markets — is subscription fatigue and price sensitivity.

Consumers are becoming more selective. The evidence: streaming subscriber growth has slowed sharply and churn has increased as prices rise and economic pressure increases. SaaS companies are facing harder renewal conversations. The era of passive subscription accumulation is giving way to more deliberate evaluation.

That’s probably healthy. Subscriptions you actively choose because they genuinely serve you are one of the better commercial models available. Subscriptions you keep by default because cancellation is inconvenient are a drain with no corresponding benefit. The difference is worth fifteen minutes with your bank statement.