The State of Newsroom TTV
Text-to-video (TTV) is changing how newsrooms produce. The TTV meant here is not generative clips conjured from a prompt, but the practice of turning already published article text into video. We cross-analyzed roughly 6,700 videos that 38 news outlets made with Videostew in the first half of 2026, combining production data with YouTube performance.
What adoption changed: dormant channels wake up
Comparing each channel's upload history before and after adoption, the change appears as a step. Upload frequency is computed from each video's publish time, so older videos do not inflate the figure. The change held regardless of size. The same step showed up at small online outlets and at national dailies with hundreds of thousands of subscribers alike.
Who uses it: still a market of a few adopters
Many outlets have tried TTV at least once, but few use it steadily. Outlets that stopped after a single month are more than half by count, yet they account for under 4% of total volume. Whether adopted via SaaS or wired in via API, trial is wide and retention is narrow.
How they use it: an outlet's character sets its video grammar
At the same size, outlets of different character make opposite videos. Length and orientation are set by an outlet's character more than its size.
How much effort it takes: low cost, high speed
How easy TTV actually is shows plainly in the production data. Tool difficulty did not vary with size. Larger outlets simply run the same tool across more teams at once. This tendency is stronger where video generation is wired into the article flow via API. In the SaaS approach, where people build videos by hand, about 22% of generated videos never reached final posting, whereas in the API approach that share was just 1%.
As it gets easier, larger outlets add more people. Large outlets split workspaces by desk, like the newsroom or the securities desk, so many reporters build videos at once in their own spaces. Small outlets usually work as one person in one space. At one regional outlet, over the 4 weeks before local elections, a single person made about 120 videos, absorbing alone an election-coverage demand that spiked from two to four a day.
More volume without losing quality
A common worry is that churning out videos through automation drags quality down. The data showed the opposite. Because the source is a real published article rather than a scene conjured from a prompt, every video has something to convey, and viewers watch to the end. That the source is an article is what sets newsroom TTV apart from ordinary generative video.
Raising volume and holding per-video quality did not conflict. One regional outlet raised output from two to four videos a day during a local election, yet median views per video rose from 536 to 927 and channel impressions grew about 2.7x. The more it published, the more the channel's total views and subscribers grew together, with each video bringing in more than one new subscriber on average. At channels that sharply increased uploads after adoption, views per video mostly rose too. Here, producing more and protecting performance pointed the same way.
How many published articles become video
The share of published articles turned into video varies widely by outlet. TTV is not yet the mainstream of editing, but closer to a supplement that extends a portion into video.
The production bottleneck is gone. Next come retention and results.
In the first half of 2026, TTV became a real production line, not a newsroom experiment. Every channel that adopted it raised output, and the change held regardless of size. Turning text into video is already fast and cheap.
Yet a few still carry the market. Volume clusters at the handful that stuck with it, and most outlets turn only a single-digit percentage of their articles into video. Where the production bottleneck cleared, the next one remains: what to turn into video, what to release, and how to connect it to results.
