Streaming looked like a simple business — collect subscriptions, spend on content, keep the difference. It has turned out to be considerably more complicated, and the shifts explain most of what subscribers have noticed.
The original model
Growth first, profit later.
The reasoning was that the winner would be whoever reached scale, since content costs are fixed regardless of audience size and marginal cost per additional subscriber is close to zero.
Which justified enormous content spending funded by debt and by investor patience, on the expectation that subscriber numbers would eventually support it.
Markets rewarded subscriber growth and largely ignored losses for several years.
What changed
Interest rates rose, which made debt-funded growth considerably more expensive.
Subscriber growth slowed in mature markets as penetration approached saturation.
And one major service reported a subscriber decline, which prompted a rapid reassessment across the sector.
Investors shifted from rewarding growth to demanding profitability, essentially within a quarter, and every service responded.
The responses
Price increases, repeatedly, across essentially every service.
Advertising tiers, which had been resisted on brand grounds and became universal quickly.
Restrictions on account sharing, which converted some non-paying viewers into subscribers and lost others.
Content spending discipline, with cancellations of expensive productions and reductions in commissioning volume.
And licensing content to competitors again, having previously withdrawn it to build exclusive libraries.
Why advertising tiers matter more than they look
The revenue per subscriber on an advertising tier can exceed that of a standard subscription, depending on how much the viewer watches.
Which inverts the assumption that ad tiers are a discount product, and it explains why services have promoted them so hard.
It also changes what content is valuable — a show that generates many viewing hours is worth more under advertising than under subscription, where only retention matters.
That shift is visible in commissioning and will become more so.
The cancellation pattern
Series being cancelled after two seasons became noticeable and has a specific logic.
Cast and crew costs rise with each season under standard contract structures.
New subscriber acquisition attributable to a returning series declines, since the audience already subscribed.
And in some cases, removing a completed series from the catalogue produced accounting benefits.
Which means the economics of a third season are worse than those of a first, regardless of quality or audience.
The bundling return
Services have begun bundling with each other and with telecoms and retail subscriptions.
Which is what cable did, and the sector has arrived back at a version of the thing it disrupted, for the same reasons — bundles reduce churn and spread acquisition costs.
Churn is the central metric now. Subscribers who join for one series and leave are expensive, and reducing that behaviour is worth more than adding new ones.
What this means for viewers
Prices will keep rising, since the sector needs to reach the profitability it deferred.
Libraries will keep changing, since licensing is now transactional again.
Fewer expensive series will be commissioned, and more will end early.
And the ad-free premium will grow, since the ad tier is where the margin is.
None of which is unusual. It is a maturing industry doing what maturing industries do, after an unusually long period of subsidised expansion.
Sport and why it changed everything
The category that broke the assumption that streaming and traditional broadcasting were separate businesses.
Live sport retains audiences that scripted content does not, because it must be watched at the time and cannot be replaced by a similar alternative.
Which makes it the most effective tool against churn available, and streaming services have bid aggressively for rights they previously avoided.
The economics are difficult — rights costs are enormous and fixed, and the audience arrives for the sport rather than for the service.
Whether it retains subscribers between seasons is the open question, and the early evidence suggests substantial cancellation when a season ends.
International production
One of the clearer structural changes.
Services commissioning locally in many countries discovered that some of those productions travelled internationally.
Which was not the original purpose — local content was for local subscribers — and it has produced genuinely global hits from production centres that previously exported little.
Local content quotas in several jurisdictions have accelerated this, requiring a proportion of catalogue and investment to be produced locally.
The result is more production in more places, which is a real change and one of the more positive consequences of the whole period.
Measurement
Services disclose viewing figures selectively and on their own definitions, which vary and change.
Independent measurement has improved in some markets and remains partial, which means most public discussion of what is popular rests on numbers the interested party chose to release.
The password sharing arithmetic
Restricting sharing was expected to lose subscribers and gained them on net in most markets.
The reason is that most affected users converted rather than leaving, generally to cheaper ad-supported tiers.
Which was a better outcome than the sector expected and has been widely copied since.