Warner Bros. Discovery EMEA: 70% Less Time Searching and Clipping Across Cartoon Network and Boomerang

Case study summary: Warner Bros. Discovery EMEA uses Imaginario AI to search and clip the Cartoon Network and Boomerang archives for YouTube and social media. Two teams in London, video marketing and post-production, search decades of kids and family programming in plain language and pull frame-accurate moments at shot level. Four years in, the teams spend 70% less time searching and clipping for social media, and ad sales have grown by more than 10% year on year.

BACKGROUND

Warner Bros. Discovery EMEA runs Cartoon Network and Boomerang across the UK and EMEA out of London. The core team we work with handles video marketing and branding, running the YouTube strategy, the channel calendars and the editorial campaigns that surround every show. We also work with WBD’s creatives in post-production, who prepare and deliver the material and sit closest to the library. The teams in France, Italy, Spain, Turkey, the Middle East and Africa use the platform as well.

For children’s media, YouTube is not a secondary marketing surface. It is the primary one, and the teams manage channels with tens of millions of subscribers between them. The main UK channels alone count more than 11 million subscribers on Cartoon Network UK and more than 8 million on Boomerang UK. That means publishing constantly, across long form, shorts and vertical, in multiple aspect ratios and territories, every week. It is a factory, not a campaign.

THE LIBRARY

We do not process the whole Warner Bros. Discovery library. We process the top kids franchises, which is where the publishing demand actually sits. That slate includes The Amazing World of Gumball, Adventure Time, Steven Universe, Apple & Onion, Ben 10, Clarence, Craig of the Creek, Looney Tunes, Regular Show, Teen Titans Go!, Scooby-Doo, Tom and Jerry and We Bare Bears, among others.

Each of those is seasons or decades of episodes, and most of it has only ever been cataloged at the title or episode level. That is fine for rights and distribution and close to useless when someone needs the eleven seconds where a specific character does a specific thing.

THE CHALLENGE

Ask anyone in either team and they will tell you the moment they need exists. They can usually describe it precisely. What they cannot do is get to it quickly. The traditional route is to guess which episode it lives in, open it and scrub. Failing that, search filenames, ask a colleague who has been there long enough to remember, or dig through legacy metadata written for a different purpose a decade ago. For a single clip that is annoying. Across a franchise slate this deep, several territories and a weekly schedule, it becomes the largest cost in the workflow.

Part of the problem is the tooling. There is a traditional MAM system in the building, but creative teams do not necessarily work with it. It serves the content operations and engineering functions and it is built for them, and like most enterprise MAM and DAM systems it was not made to understand video, find moments inside it without tags and metadata, or be intuitively simple for creative teams. Imaginario leapfrogs the MAM entirely. For this workflow it is not needed, and the content ultimately lives in cloud storage services like AWS S3, which is exactly where we plug in.

It also quietly caps ambition. When finding a moment is expensive, teams reuse the same well-known moments, because those are the ones they can locate. The long tail of the catalogue, which is where the fresher material sits, goes untouched.

THE SOLUTION

We indexed the franchise libraries using our proprietary AI systems across visuals, speech, sound and people, without relying on manual tags or labels. Teams then search in plain language and get results at shot level, with frame-accurate boundaries, rather than a full episode to scrub through. Over four years the work has settled into three patterns.

Single cuts from long-form episodes. The most common job. Take a full episode and pull the one self-contained moment that works as a standalone post. A gag, a stunt, a reveal, a line that travels on its own. The episode runs twenty-two minutes and the post runs forty seconds, and almost all of the effort used to go into locating the forty seconds rather than cutting them. The same moment rarely ships once. It gets reframed and reformatted for YouTube, Shorts and the vertical platforms, so one good find feeds every platform on the calendar.

Seasonal compilations across episodes and shows. Halloween, Christmas, back to school and the rest of the calendar all need packages that gather the right moments from across different episodes, different series and different show brands. This was close to impossible to do well before, because nobody could survey the whole slate for a theme. The realistic approach was to use the episodes the team already remembered. Searching by theme, mood, setting and action means a Halloween compilation can now draw from titles nobody had opened in years.

Character identification. This is the capability that mattered most, and it is what separates children’s animation from most other video work. These brands are built on characters. Almost every brief, editorial or commercial, is phrased in terms of a character rather than an episode. Standard face recognition is trained on human faces and does not help here. Identifying specific animated characters, and then finding every moment they appear in across the slate, is what makes the rest of the workflow usable. Without it, a search for a character returns episode titles. With it, it returns moments.

Finding the moment is half the job. The teams finish it in Clip Studio, where they finesse the in and out points, add branding and add subtitles. For bigger packages they group moments into collections and export them to Adobe Premiere, which Imaginario supports natively, so an editor picks up a sequence rather than a folder of loose files.

None of this works without the plumbing. Connecting directly to AWS S3, where much of the material lives, meant the teams never had to move or duplicate assets to make them searchable. At the other end, results go straight into Premiere. Imaginario sits as a horizontal layer on top of the storage and editing infrastructure that was already there, so nothing had to be replaced.

Language matters just as much across EMEA. The platform supports more than 100 languages and dialects, which helps the teams in other territories enormously with captioning, and the search engine natively supports more than 17 languages, so a producer in Madrid or Istanbul can search the same library in their own words.

The saving is not confined to one desk. Assistant editors, producers and marketers all spend a large part of their week locating material before any creative decision gets made, and all three groups got that time back. Output went up, because the same headcount ships more per week and publishing frequency is one of the strongest levers on YouTube views. Selection improved, because when searching is cheap teams explore, and older titles and secondary show brands come back into rotation.

The audience numbers tell the same story. When Cartoon Network UK celebrated reaching 10 million YouTube subscribers in early 2024, the team had already been working with Imaginario for a year and a half. Since then the channel has added close to 2 million more, growing to over 11.8 million subscribers, an increase of almost 18% in about two and a half years. That growth has many drivers, and we would not claim sole credit, but publishing more, fresher moments every week is one of them.

A CLEAR RETURN ON INVESTMENT

Four years of daily use have produced tangible benefits for Warner Bros. Discovery EMEA, including:

Ad sales growth always has several drivers and we would not claim sole credit, but faster, character-accurate packaging is a direct input into both the volume of sellable inventory and the conversation about what it is worth.

KEY FACTS

• Company: Warner Bros. Discovery EMEA, kids and family brands, running Cartoon Network and Boomerang out of London
• Teams using Imaginario AI: video marketing and branding, and post-production in London, plus the teams in France, Italy, Spain, Turkey, the Middle East and Africa
• Users: assistant editors, producers and marketers
• Use case: plain-language search and shot-level clipping across franchise libraries, seasonal compilations, animated character identification, with content staying in AWS S3 and results exporting to Premiere
• Library: top kids franchises including The Amazing World of Gumball, Adventure Time, Steven Universe, Ben 10, Looney Tunes, Regular Show, Teen Titans Go!, Scooby-Doo, Tom and Jerry and We Bare Bears
• Working with Imaginario: 4 years
• Results: 70% less time searching and clipping for social, over 10% year on year growth in ad sales

If your team is sitting on a library it cannot fully see, we should talk.

Read next: How Cineverse searches its 60,000-title library and cut video search and clipping time by 75%.

FREQUENTLY ASKED QUESTIONS