Streaming Growth Fell Into Single Digits, So HBO Max Built a TikTok

Content is hard. Whether you are making it, selling it, marketing it or any of the other myriad roles we have in this industry, what we do isn't easy. Fun, joyous, exciting, sure. Easy... not so much.

Each show, each episode, is a beautiful, idiosyncratic thing, born from a community and then growing, or dying, from there. Which is a long way of saying that as a child, when I stared (and starred, and stared) at the TV set, I knew I wanted to help make these beautiful things. And now I have, and it is amazing and there is so much content for us to watch. So much content. Like a stunning amount of content. To the point where most nights I find myself scrolling tiles across five different streaming apps, all of which I pay for, and then, in frustration, hopping to YouTube to let the algorithm take me on a ride.

Last Tuesday, Warner Bros. Discovery (someday I'll write about the merger but in the meantime I just want to say to anyone reading this who works there or at Paramount, I see you, I feel you and I've been there) shipped a feature designed for exactly that evening moment (rolling out first to select iOS users in the U.S.). HBO Max Shorts is a vertical, scrollable feed of clips, trailers and bonus scenes, sitting behind a new icon at the bottom of the app, personalized to your watch history. An in-house model scores scenes across the catalog for how likely a clip is to hook someone who has never seen the show, and editors pick the finalists. From there it plays out fairly easily like the clip - tap and start the episode. They also included a conversational search tool along with it to help further drive people through the experience.

And HBO Max was last (which is a shame to write, because it's been such a fun season of House of the Dragon).

Tubi had Scenes running back in November 2024, and ESPN followed with vertical highlights in August 2025. Then the bigger players moved in - Disney+ shipped Verts in March, Peacock had been rolling pieces of a vertical feed out since early 2025 with a dedicated destination still arriving that summer, Paramount+ added its own version the same spring, Netflix launched Clips on April 30, and Prime Video expanded its version beyond NBA highlights in May. So now we've got eight services from seven companies that tend not to agree on a lot of things, all landing on a very similar product concept in under two years.

So you gotta ask - why? Why now and what are they looking to accomplish (and is it the same goal for each of them?).

The reductive explanation is this solves for attention spans and churn. Everyone is doomscrolling, subscribers are fleeing, and streamers are building mini panic rooms that look (perhaps not surprisingly) a lot like TikTok and may or may not contain Jodie Foster (I stand by this joke).

It's a fun story to write, unfortunately it is not the story that the data supports (so disappointing when this happens). Antenna, which measures U.S. subscription behavior at the transaction level, puts the weighted average monthly churn rate for premium streaming at a steady 4.6% across 2025, and describes the year as stabilization rather than crisis (go teams!). So if subscribers are not fleeing, what is everyone doing here?

Antenna estimates subscribers across the entire premium streaming category, every service combined, grew just 7% in 2025, down from 12% the year before, the first single-digit year in the category's history.

That... is not great for businesses that run on recurring subscription revenue.

When acquisition is driving growth, the job is to convert a new person as fast as possible and then get the hell out of the way. Discovery failure is expensive but survivable, because new customers keep arriving to replace whoever falls off. Once acquisition stops carrying your growth, the subscribers you already have become THE thing, and the little annoyances they used to shrug off start costing real money (especially in a world where it is now exceptionally easy to cancel a subscription). In that world, a catalog that people don't dig into is a very expensive thing with very little actual current value.

If you spent the last decade buying content, the math here is... challenging. The libraries these companies built are enormous, expensive, and largely invisible to the subscribers funding them (although I am binging Survivor right now from episode one, so...), because thousands of hours sit behind a homepage that surfaces maybe forty tiles, many of which feel like the same forty titles you saw last week.

The vertical feed is the cheapest available answer, because it requires no new content spend, no new licensing, and no production cycle. It repackages assets already paid for and aims them at the only audience left to grow, the one still on the streamer.

Wild stat, right?

Antenna found that between September 2023 and August 2024, roughly 34% of all gross additions across premium streaming were resubscribers, people rejoining a service they had canceled within the previous twelve months. So a third of the incoming class already knows the product, and the last thing they did with that knowledge was cancel.

To be fair, what we don't have a sense of is how much of that is driven by seasonal content. So when House of the Dragon ends and I drop off (I'm not, actually, HBO has me in terms of value of content delivered), that is its own decision and challenge.

The whole thing becomes a bit of a nightmare to handle. The person signing up this month may be someone who canceled back in March. They know the library, they know the homepage, and they decided to get out a few months back. Now one title has them resubscribed and the streamer gets the length of whatever that was (most likely a 6-8 episode series) to convince that person to not cancel again. This is hard work, one that a homepage usually can't solve because the homepage assumes patience and rewards familiarity, and many viewers today... aren't necessarily good at that whole patience concept. So the new new thing is trying to find the best ninety seconds the catalog has to offer, served before they even start thinking about setting a reminder to cancel in two months.

The feed is a re-onboarding machine for an audience that keeps leaving and coming back and the streamers are building this because new customers aren't there at the level they need or want.

For a generation, the brand itself was the thing. You weren't evaluating individual shows, you just knew that every Sunday night HBO would have a big cool new series (along with some movies and standup specials) and they would take care of you. Premium product versus everything else available.

Now things have shifted - the brand still sells the subscription but, if this works, a feed now decides what you actually watch, making a fresh case every time you open the app.

Podcasters have spent years with this, being told that clips are marketing, a promotional layer that lives on social platforms and points back toward the real product. The streamers just made a very different call - they built the feed inside their system, gave it permanent navigation, and pointed it at their own catalog, because this is all about discovery and retention. For years the discovery moment happened on other companies' apps, on their algorithms and their terms, and the streamers decided they want that moment happening somewhere they own.

Which gets to the bigger question - if short-form is the layer where people decide what to spend their time on, and you are running that layer on platforms you do not control, then you're relying on another model to determine where (and who) that content is served to. I wrote a while back about the cost of renting someone else's fandom. This is a similar take on the platform side.

Every one of these feeds is designed to end in long-form playback, which is what separates them from the social channels you'd expect to find them on. The feed exists to start an episode, and that is it. Nobody building these wants you spending three hours scrolling clips. They want the one clip that tips you into pressing play, because once acquisition stops carrying growth, conversion inside the app is what keeps the customer list from atrophying.

Thanks for reading and if this was useful, feel free to forward it to someone who should be thinking about this. Join a growing group of operators, creators, and strategists focused on how media actually works.