The Subscription Economy Beyond Streaming

It is worth being precise about the customer's side of the exchange, because most subscription business cases describe the operator's side in detail and the customer's side in one line.

• Convenience: removing a recurring task from their life.
• Access: use of something they do not want to own, maintain or dispose of.
• Predictability: a known monthly cost instead of a lumpy and uncertain one.
• Lower upfront commitment: capital preserved for other uses.
• Flexibility: the ability to change, pause or exit as circumstances change.
• Personalisation: a service that improves as it learns about them.
• Ongoing service: the assurance that problems are somebody else's to solve.
• Reduced ownership complexity: no resale, no depreciation risk, no administration.

A proposition that delivers none of these, but charges monthly, is not a subscription business. It is a financing arrangement, and it will be competed away by whoever offers a cheaper one.

Market sizing in this category should be handled with care, because the definitions are not consistent. Research estimates have valued the global subscription economy at approximately 492 billion dollars in 2024, with forecasts of roughly 1.5 trillion dollars by 2033 at a compound annual growth rate near 13%. Other work, including Juniper Research, has put the market at around 722 billion dollars in 2025 rising towards 1.2 trillion dollars by 2030.

These figures are not contradictory so much as incommensurable. They include different categories, some counting only consumer digital subscriptions, others including subscription commerce, business software and services. Both are forecasts, which means both embed assumptions about adoption that may not hold. The useful conclusion is directional: subscription is expanding beyond media into categories with physical fulfilment and real assets, and the operators entering those categories will meet economics that streaming never had to solve.

One of the most important lessons I learned while building Carasti was that asset-heavy subscription is a utilisation business wearing the clothes of a consumer brand.

Utilisation is the master variable. An idle asset does not merely fail to earn, it continues to depreciate and to consume financing, insurance and storage. A model that assumes ninety per cent utilisation and delivers seventy is not ten per cent worse, it is frequently unprofitable.

Residual value is an exposure you carry whether you acknowledge it or not. In practice you are running two businesses simultaneously: a service business and an asset trading business. The second one can quietly consume the profits of the first if the timing of disposals is wrong.

Acquisition cost interacts with churn in a way that is easy to underestimate. In monthly-cancellable models, the payback period is not a static number, it is a distribution, and the tail matters. A cohort that averages nine months of tenure can still be unprofitable if a large share leaves in month two.

Pricing must reflect the cost of flexibility. Flexibility is the product, and it is expensive to supply, because it transfers risk from the customer to the operator. Charging the same for a twelve-month commitment and a rolling monthly arrangement is a subsidy that will be arbitraged immediately.

Service delivery is where retention is decided. Customers forgive price. They do not forgive being without the thing they are paying for. In an operationally intensive subscription, the operations team is the retention team, whatever the organisational chart says.

And supply and demand matching is a continuous problem rather than a planning exercise. Demand varies by segment, season and location. Supply is committed months in advance and cannot be reallocated quickly. Most of the operating skill sits in that mismatch.

In an asset-heavy subscription business, the balance sheet has an opinion about your marketing strategy, and it is usually right.

A subscription model can be entirely unattractive despite growing recurring revenue. The warning signs are consistent and largely mechanical.

• Poor gross margin, which no growth rate will repair.
• Expensive acquisition relative to expected lifetime contribution.
• High churn, which converts acquisition spend into a treadmill.
• Weak differentiation, forcing price competition in a category with high fixed costs.
• High servicing cost per subscriber, particularly where physical fulfilment is involved.
• Underutilised assets, in models where utilisation drives contribution.
• Failed payments, which in some markets remove a meaningful share of billed revenue before it is ever collected.
• Excessive discounting at acquisition, which selects for price-sensitive customers who churn at the first increase.

Any two of these together will usually make a subscription business unviable regardless of top-line growth, and the reported annual recurring revenue will look healthy right up until the cohort curves are examined.

There are two ways to retain a subscriber. One is to make leaving difficult. The other is to make staying obviously worthwhile. Both show up identically in this month's churn number and diverge completely over three years.

Retention built on value comes from four sources: the product genuinely solving a recurring need, habit formed through regular useful interaction, service quality that removes anxiety, and a proposition that improves with tenure. Retention built on friction comes from difficult cancellation, obscured terms and inertia. The second is being regulated out of existence in several jurisdictions, and it deserves to be.

My practical view is that a business should be able to state, in one sentence, why a customer would choose to stay next month if cancelling took thirty seconds. If that sentence is hard to write, the retention number is borrowed rather than earned.

Subscription fatigue is real and increasingly measurable. Households are auditing recurring commitments, and any subscription that is not clearly used will be cancelled during that audit. New entrants should assume they are competing for a slot rather than creating one.

Cancellation friction and hidden charges are attracting regulatory attention in several markets, and rightly so. Business models that depend on customers forgetting they are subscribed carry regulatory, reputational and valuation risk.

Affordability matters. Converting a large one-off purchase into a recurring commitment can improve access, and it can also lock households into obligations that become difficult during income shocks. Responsible operators build pause and downgrade paths before they are demanded.

Asset risk deserves particular caution. In asset-heavy models, a downturn in residual values can eliminate several years of service margin, and that exposure is often invisible in reported subscriber growth.

There is a legitimate counterargument that ownership remains superior in many categories: for high-utilisation, long-lived, emotionally significant goods, subscribing is simply more expensive over time, and customers eventually work that out. Subscription is not a universal model, and the categories where it is genuinely better are narrower than the enthusiasm suggests.