Great ResignationClimate ChangeLeadershipInflationUkraine Invasion

AI Is About to Make Video Ads a Lot Less Annoying

June 26, 2018, 1:49 PM UTC

No one likes having their streaming video experience interrupted by pre-roll advertising that doesn’t appropriately match the content and video duration; it can be inherently disruptive. In fact, a 2017 report from IPG Mediabrands’ Media Lab found that 65% of pre-roll ads are skipped. This is challenging for streaming video platforms and content publishers that want to provide an optimal viewer experience, but need to monetize their content too.

So what’s an alternative?

While ad-free subscription models have worked for some platforms—notably Amazon (AMZN) Prime and Netflix—they’re hardly a panacea. Three years after its launch, for example, YouTube’s subscription service YouTube Red still only accounts for 7% of the company’s revenue. Smaller subscription services are beset with high customer churn as viewers skip from free trial to free trial. “Subscription fatigue”—the particular type of exhaustion customers experience from paying too many fees to too many platforms—is rife.

Luckily for platforms, creators, and users alike, artificial intelligence (AI) presents an answer—and that is to move advertising inside TV shows rather than just around them. Product placement is nearly as old as television itself. However, AI may soon put a new spin on that time-tested strategy, making it scalable, measurable, and personalized in the way that pre-roll ads are today.

If you’ve used an augmented reality (AR) app like Pokémon Go, you’ve already experienced the power of surface detection technology. This AI-powered tool is what allows the app to place a Pikachu on the real-life table you see through your camera lens, rather than leaving them floating in space. Earlier this year, Apple (AAPL) announced it was enhancing its AR toolkit for developers to enable detection of vertical surfaces, too. That means AR apps can make posters appear on walls or paint them with splashy colors and designs, and already, the technology has been used to create more interactive advertising experiences.

That same technology will soon be used to automate the placement of unobtrusive ads in the background of TV shows and movies. Studios won’t have to change anything about how they film or edit their content; however, when the final video is delivered to the viewer on a smart TV or other connected device, ad servers will use surface detection technology to insert products or advertising directly into specific scenes. For instance, a server could put a bottle of a particular soda on a character’s kitchen table or fill in a billboard outside her window with a subtle ad. What’s more, all of those insertions would be personalized, so no two viewers see the same exact thing. A music fan might see an ad for a new hip-hop album on that billboard, while someone who’s recently searched for auto dealerships in her area might see a car ad. If successful, such insertions could eventually replace interruptive commercial breaks.

This type of programmatic product placement is likely to start hitting your smart TV screen in a year or two. The technology is there—it’s just a question of advertisers and content creators getting on board. Right now, many of them are still hesitant to cede control over ad placement to AI algorithms. However, with mounting pressure to make online content more profitable, it’s only a matter of time until the most forward-thinking among them begin experimenting with these types of solutions. When that ultimately happens—and leads to widespread adoption—it’ll be a big win for users and companies alike, bringing a more relevant ad experience and greater monetization opportunities for publishers.

The key to the latter will be tight integration of in-episode ads with other technology in the home. That music fan should be able to click the billboard on their smart TV screen—or just point their smartphone at it—to see options for buying that album on an online streaming service like Spotify (SPOT). Studies show that most viewers are already using another device while watching TV, so why not involve that device directly in the experience?


Plus, with the upcoming arrival of 5G wireless networks—which will have faster download speeds than most broadband—consumers may watch more high-definition video on their phones and tablets anyway. Smart speakers and digital assistants could get in on the act, too, integrating tightly with smart TVs and devices to smooth each viewer’s path from viewing an ad to making a purchase.

Streaming video is on track to make up 82% of all consumer IP traffic by 2021. Increasingly, it’s the dominant form of media online. That means that a bad user experience for video—riddled with annoying, glitchy, or just plain obtrusive ads—is a bad user experience for the Internet, full stop. By making ads more streamlined and subtle, programmatic product placement and other AI-driven technologies won’t just help publishers monetize content more effectively. They’ll make the Internet a better place for companies and users alike.

Ralf Jacob is head of digital media services at Oath.