This isn’t the industry’s first time making this assumption. When older generations were young, the expectation was that they’d start reading newspapers as they aged, the way generations before them had. When that pattern weakened, the assumption shifted again: audiences would simply grow into consuming news online instead. For a decade, the same logic has been applied to social and video-first audiences: give it time, the thinking goes, and today’s 22-year-old scrolling TikTok would settle into the habits of a 45-year-old checking the homepage every morning.
The Reuters Institute’s 2026 Digital News Report is based on nearly 100,000 interviews across 48 markets, and within that sample, the pattern for young adults is hard to ignore.
Globally, more than half of 18-24 year olds now say social media, video networks or AI are their main source of news. And there’s little evidence they’ll grow out of it: 56% of 18-24 year olds who hadn’t read a newspaper in the past week said they’ve never regularly read one at all, compared with just 21% who say the same about television. People aren’t delaying old habits. They’re forming different ones from the start.
This is usually framed as a strategy or brand problem: how do you sound authentic on TikTok, or how do you build trust with an audience that’s never subscribed to anything? Those questions matter. But underneath them sits a quieter, more structural one that gets less attention: can your production workflow actually keep up with where this audience already is?
The shift is bigger than "post more on social"
The pattern is familiar. What’s different this time is the scale, two findings from the same report make that clear.
First, the age gradient is steep and getting steeper. 52% of 18-24 year olds say social media, video networks or AI are their main source of news which is 32 percentage points ahead of any other source. Among 25-34s, that figure is already down to 44%. The younger the reader, the sharper the shift.
Second, video habits follow the same pattern. Almost four in ten 18 to 24 year olds use TikTok weekly for news videos, while 36% use Instagram. Among users aged 55 and older, the figures fall to just 12% and 15%, respectively. Text-first publishing, still the default architecture behind most newsroom CMS platforms, is no longer matched to where this audience spends its attention.
Together, these two findings shift the question. It’s less about whether a newsroom should have a TikTok strategy, and more about whether a single piece of reporting has a realistic path to becoming a short vertical video, a text explainer, and a long-form web article at the speed a platform-first news cycle demands – without three separate teams rebuilding it three separate times.
For much of the industry, that’s still a stretch. Not because the editorial appetite isn’t there, but because a lot of newsroom content architecture was built before video-first, platform-native formats were the majority rather than the exception. In that kind of setup, video tends to become a separate production with its own team, its own timeline and disconnected from the reporting that generated it rather than a natural extension of the story.
There’s a difficult question underneath all of this that this piece isn’t trying to answer: distribution moving to platforms a publisher doesn’t own is only half the threat; the other is a shrinking share of any audience willing to pay for news at all. No production architecture fixes either one. What it can do is make sure that when a newsroom does land on a viable model, the content itself isn’t the bottleneck standing in the way. It starts with how a story is built in the first place.
Where the story-first model actually earns its keep
A story-centric content architecture is the practical answer to that gap, not a generic efficiency pitch, but a direct response to how this audience consumes news.
In Cue Content Store, a story exists as one entity with multiple variants, rather than as separate builds per channel. That matters here because it means a reporting team’s core work: the facts, the sourcing, the structure, doesn’t have to be re-assembled from scratch to exist as a 30-second vertical clip instead of an 800-word article. The core story stays connected to a fast-turnaround video variant the same way it stays connected to a newsletter version or a print layout.
Cue Dam plays a similar role on the asset side: as video becomes the most-consumed format rather than a supplementary one, the raw footage, b-roll, and graphics behind a story need to be centrally searchable and reusable, not siloed in whichever team happened to shoot them.
Swiss broadcaster RSI illustrates the same shift from a different starting point. Rather than moving from text into video, RSI was a broadcast-first operation looking to accelerate digital publishing. By connecting its media asset management system directly into the Cue platform, RSI made it possible for video segments from its main TV channels to be published online within seconds of airing, and for information from a live broadcast to populate content templates automatically as the show is happening.
The underlying shift is the same one the data points to: treating video as a connected part of the story from the start, rather than a separate deliverable handled after the fact. None of this replaces the editorial and creative judgment that a platform-native audience demands: the tone, creator-style delivery, and platform-specific pacing that separate content that works on TikTok from content that merely exists there. That’s a people and strategy question, and it’s a different one. What the underlying architecture can do is remove the production bottleneck that holds back that strategy from shipping at the pace the audience demands.
Where this leaves newsrooms
This is not about chasing every platform. It’s about ensuring the production model can support the audience strategy. A newsroom that solves the brand and tone question will still be a step behind if a single story can’t move from reporting to short-form video without being rebuilt from scratch each time.
Curious what a story-first architecture would need to look like for your own multi-format output? Get in touch with the Stibo DX team to talk through it.