YouTube used its 23 September Made on YouTube post to preview two Studio features that both choose between creative options, in different ways. Video-cut A/B testing will let creators “test up to three video cuts to see which hook performs best,” carrying the site’s existing title-and-thumbnail contest into the edit itself. Dynamic thumbnails, announced in the same post, recommend the best of three thumbnail options to different segments of the audience, rather than one image for every viewer.
Cut Tests Extend an Existing Contest
Today’s Test & Compare tool already runs a similar contest for titles and thumbnails. Creators upload up to three versions, YouTube shows them to real viewers, and within two weeks the tool returns a “Winner,” “Performed same” or “Inconclusive” call. YouTube ranks that contest by watch time, saying the metric “will best support creators’ growth,” rather than click-through rate. The tool works only in desktop Studio, needs advanced features turned on, and, per YouTube’s help page, “is not available for Shorts.”
Video-cut A/B testing borrows the same shape and moves it into the edit itself. YouTube’s blog post says creators will “test up to three video cuts to see which hook performs best.” The Creator Insider demo video published two days later, as transcribed by PPC Land, describes YouTube running the cuts “before audiences” and “giving you a breakdown of watchtime retention curves.” YouTube’s own tally for the existing contest is large: “creators have run more than 40 million experiments to find the perfect titles and thumbnails for their videos since the feature officially launched in 2024.”
What are YouTube dynamic thumbnails?
Dynamic thumbnails are a YouTube Studio feature announced on 23 September that lets a creator supply three thumbnail options and have YouTube “recommend the best of three options to different segments of your audience,” rather than showing one image to every viewer. YouTube hasn’t said whether dynamic thumbnails return any verdict like Test & Compare’s Winner, Performed same or Inconclusive. Nor has it said how those audience segments are formed, or whether a creator can see which image reached which segment.
Timing and the Shorts Conflict
YouTube’s blog gives no date: the feature is “coming soon.” TechCrunch puts a date on it anyway, reporting that “from 2027, this feature will let creators upload and test three different cuts for Shorts,” a claim that sits awkwardly next to YouTube’s own rule that today’s A/B testing “is not available for Shorts.” PPC Land adds a second wrinkle: the Creator Insider demo narration describes cuts for “shorts or long form,” while the video’s written description mentions Shorts without settling whether long-form video qualifies too. The two accounts don’t reconcile, and nothing in the primary sources says which format ships first.
What YouTube Hasn’t Said About Sponsored Video
PPC Land lists a set of questions the announcement leaves open, all specific to brand and sponsored video: whether a video carrying paid-promotion disclosure can enter a cut test, whether every cut must keep the brand integration, whether the disclosure label follows each variant, and which cut runs when a brand uses the video as a Google Ads ad. YouTube hasn’t said.
YouTube’s blog post also extends automation past manual testing. YouTube says “Ask Studio can even run in the background to review your older videos to suggest and test refreshed thumbnails, helping classic content reach new audiences.” TechCrunch describes the same capability with a condition attached: “If a creator allows it, Ask Studio will be able to autonomously monitor thumbnail performance and create tests to refine older videos.” Neither source says whether that automated review will extend to cut testing once it ships.
The Read-Out Is the Missing Piece
For a channel that also buys media, the difference between the two features matters. If the demo holds, a cut test hands back retention curves, a finding a marketer can carry into the next edit or into a paid cutdown. YouTube hasn’t described what dynamic thumbnails report back: nothing in the announcement mentions a per-segment breakdown for the creator. Pinterest left a similar gap in its Visual Search Ads announcement, which named no reporting line for the new format. Snapchat went the other way this month with its third-party web optimization beta, which optimizes toward the third-party purchase outcomes advertisers already use to judge campaigns. On YouTube, the open question is the read-out, not the test itself.
