YouTube first spoke about pause ads last year when it started trialing them in select regions. At the time, the company said that when you pause a video, it will shrink, and an ad will appear next to it.

Example:

“In Q1, we saw strong traction from the introduction of a pause ads pilot on connected TVs, a new non-interruptive ad format that appears when users pause their organic content,” Schindler noted. He went on to share that YouTube’s pause ads are “driving strong brand lift results” and “are commanding premium pricing from advertisers.”

Schindler didn’t share any timelines for when pause ads will start appearing on YouTube, but we know they’ll first roll out on smart TVs. The nature of these ads, including their duration, skippability, and more is still unclear. We also don’t know if Google plans to introduce these ads on YouTube’s mobile apps.

  • @ours@lemmy.world
    link
    fedilink
    English
    98 months ago

    Yeah, they could put the ads in the same stream but it would be too costly or inflexible. Ads have to be targeted to the specific market or even user so that would kill their advantage and turn them into generic TV ads.

    • @diffusive@lemmy.world
      link
      fedilink
      English
      48 months ago

      If you don’t have to reencode but only concatenate the streams it can be done for your request specifically because it’s not meaningfully more expensive than just serving the content

      • @lorkano@lemmy.world
        link
        fedilink
        English
        1
        edit-2
        8 months ago

        What stops you with skipping past this part of the stream? It would be more effective on livestreams, not videos. Anyway eventually people will get so pissed and every popular video could go through a server first and remove ads part via AI. There is no end to adblock war

        • @diffusive@lemmy.world
          link
          fedilink
          English
          28 months ago

          The position and length can be randomized, for sure you can throw ai at it… but ai is stupidly expensive… companies that offer AI systems are all operating that business at loss… and no big company would engage in something like this… distilled model on client? Sure… but who trains it? My point is… things can get worse and worse