For years, sharing a streaming login was normal and tolerated. Services knew it was happening at scale and said little about it. Some executives publicly described it as harmless, even beneficial.
Then enforcement arrived, subscriber numbers went up, and the earlier tolerance was retroactively explained as a growth strategy. Both the tolerance and the crackdown were rational at their respective moments, and the sequence reveals something about the business.
Why it was allowed
During the growth phase, the metric that mattered was subscriber count and market share. Every service was competing to become the default, and defaults are established through ubiquity.
Shared logins served that. A person watching on somebody else's account is a person forming a habit, learning the interface, developing preferences. When circumstances change — they move out, the relationship ends, the sharer cancels — they convert.
Sharing also functioned as free marketing. Word of mouth about a show is more effective from someone who's actually watched it.
So tolerating it was a customer acquisition cost with an unusually good return, and it was described publicly in warm terms about families and community, which was somewhat generous framing for a business decision.
Why it stopped
The market saturated. Once nearly everyone who might subscribe already has, growth has to come from somewhere else, and the obvious pool is people already watching without paying.
At the same time, investor attention shifted from subscriber growth to profitability. That changes which metrics matter and makes revenue per viewer more important than reach.
Enforcement then became straightforwardly attractive: a large population of engaged users, already habituated, who could be converted with a policy change rather than a marketing spend.
What the numbers showed
The commercial results were, by most accounts, better than sceptics expected. Conversion rates from shared users to paying accounts were substantial, and cancellations among existing subscribers were smaller than feared.
Which tells you something. The threat of losing access to something you watch regularly is a strong motivator, and the alternative — piracy, or simply not watching — was less attractive than a subscription for a large share of affected people.
It also revealed how large the shared-viewing population actually was. The gap between reported subscribers and actual viewers had been considerable, which has implications for how audience figures were being interpreted all along.
The measurement question
This is the part I find most interesting. For years, industry and press discussion of streaming reach used subscriber numbers as a proxy for audience.
If a meaningful proportion of viewing was happening on shared accounts, then those figures understated reach — while overstating the paying customer base relative to how many households were being served.
Advertising-supported tiers make this sharper. An advertiser buying against a service's audience needs to know how many people are actually watching, not how many accounts exist. Household-level and profile-level measurement has become considerably more important, and the crackdowns have improved the underlying data as a side effect.
The definitional problem
What crackdowns exposed is that "household" is a poorly defined concept in practice.
A student at university. A partner working away during the week. A parent in care. A child at boarding school. Someone travelling for months. Two people who split their time between two homes.
All of these are plausibly one household and are difficult to distinguish technically from actual sharing between separate homes. The enforcement mechanisms — device location, network association, periodic verification — approximate this and produce false positives.
The customer service burden of those false positives is significant, and it's the part of the policy that generated most of the genuine anger. People with entirely legitimate arrangements being locked out of something they pay for is a bad experience regardless of the underlying policy's merits.
What it means going forward
The broader pattern is worth naming: the generous phase of a platform business is a phase, not a character trait.
Low prices, permissive terms and tolerance of edge-case behaviour are growth instruments. Once growth is over, those instruments get withdrawn, because their purpose has been served.
This has happened in ride-hailing, food delivery, cloud storage, and now streaming. The sequence is consistent enough to be predictive.
Which suggests a reasonable posture as a consumer: enjoy the generous phase, don't build anything important on the assumption it continues, and don't be surprised when it ends. That's not cynicism about any particular company. It's just what the funding model produces.
The bundling response
One consequence that has become clearer over time: services responded to saturation not only by enforcing account limits but by bundling — packaging multiple services together, or attaching streaming to mobile contracts and retail memberships.
That is a familiar move from cable television, which is somewhat the point. A market that started by unbundling has reassembled into bundles, because bundling reduces churn and churn is the metric that matters once growth stops.
For consumers it means the effective choice is narrowing again. The promise was picking exactly what you wanted; the practice is increasingly picking a package that includes things you did not choose, at a price that is lower than buying separately and higher than what you actually use.