35% of CTV Genre Labels Are Wrong, Wurl Analysis Finds

Only 11% of CTV ad requests carry a usable show name, Wurl found, and 35% of genre labels don't match the program — a standardization gap threatening content targeting.

Wire notes

  • Only 11% of CTV ad requests contained a usable show name, per Wurl's August 2026 analysis.
  • 35% of genre labels evaluated didn't match Wurl's classification of the program.
  • Wurl tallied more than 3,600 unique genre labels in a two-hour prime-time window.
  • IAB Tech Lab's Content Taxonomy, updated for CTV in June 2022, has only 704 categories.
  • Instagram's Adam Mosseri explained why the platform runs no creator revenue-sharing program.
Future of TV Briefing: CTV advertising has a genre-label problem
PhotoFuture of TV Briefing: CTV advertising has a genre-label problem — Oscar Nord / unsplash

Only 11% of connected TV (CTV, ad-supported streaming delivered over internet-connected sets) ad requests carried a usable show name, according to an analysis by streaming tech provider Wurl. Genre labels are far more common — 93% of requests included one — but they cannot be trusted: 35% of requests checked against Wurl's own classification system carried a genre label that did not match the actual program.

The analysis covered a two-hour window during a single prime-time evening in August 2026, across Wurl-powered channels on LG Channels, Pluto TV, Samsung TV Plus and Tubi. It found 3,600 different genre labels in circulation during that window alone. The root cause, according to Wurl CEO Dave Bernath, is a lack of standardization in how content owners label their libraries.

"As a guy who has a legacy TV background – I come from the cable industry – historically, the metadata around content libraries is pretty inconsistent," Bernath said.

How bad is the labeling problem?

Two failures compound each other. First, labels are often too broad: of the 1.09 billion ad requests carrying a genre label, 22% had what Wurl considered vague designations such as "entertainment" or "general variety." Second, categorization is often simply wrong. Digiday points to "The Bear": most viewers would call it a drama, particularly in its two most recent seasons, yet FX submits it in the comedy category for awards, alongside ABC's unambiguously comedic "Abbott Elementary."

Content owners are responsible for appending labels to bid requests, and Bernath attributes their frequent misses to what he calls a "historical metadata hangover" — large, aging libraries of shows and movies originally categorized decades ago and never reviewed to see if the labels still hold.

This confusion threatens broader industry efforts to pass content signals. Over the summer, Comcast-owned FreeWheel added show-level reporting for advertisers on its platform, but it does not actively enforce common show-naming conventions between buyers and sellers — a gap that could nullify attempts to match programs and produce accurate reports.

Doesn't a standard already exist?

Yes. The IAB Tech Lab updated its Content Taxonomy (a shared framework for describing ad content) in June 2022 with categories suited to CTV, and refreshed it two years ago to add genres such as "Thriller." Even so, the taxonomy contains only 704 categories — one-fifth the number of unique genre labels Wurl observed in its two-hour sample. Wurl itself uses the IAB taxonomy as the foundation of its own system.

"We're using our machine learning, our content intelligence to analyze these streams and understand the metadata, the [electronic programming guide], to try to coalesce around the IAB standard genres specifically," Bernath said. Why other players have not followed suit, he did not say — and the industry has yet to answer.

What else happened this week?

Instagram head Adam Mosseri explained why the platform runs no creator revenue-sharing program: "Creators don't produce a lot more when we pay compared to what they already create. And what we do find is often, not always, the incremental content they create is not as high quality as what they would have created or what they were already creating for free."

By the numbers:

  • 48 million: Peacock paid subscribers at the end of Q3 2026.
  • 15.8%: YouTube's share of U.S. TV-screen watch time in August 2026.
  • $12 million: funding raised by creator-led news company Newpress.
  • 48%: expected share of midterm election ad spending going to broadcast TV.
  • -4.8%: year-over-year decline in LA-area film-and-TV shooting days in Q3 2026.

Elsewhere in coverage: midterm political spending — $8.2 billion of an expected $12 billion heading to TV or CTV, per AdImpact — has inflated prices for regional advertisers; CNN International remains on course to launch AI-agent ad sales in Q1 2027; and media buyers are charting NBA strategies as sponsorship investment rises.

In reading: Instagram is reviving long-form horizontal video (up to 90 minutes) under the name Features, with no revenue share, per The Hollywood Reporter; Netflix co-CEO Ted Sarandos will refocus on premium film and TV after ousting Bela Bajaria, per Bloomberg; and FIFA has opened the U.S. rights market for the 2030 World Cup, targeting bids by Nov. 19, with 2034 rights also on the table, per CNBC.

Until content owners adopt a shared taxonomy — or at least audit their legacy metadata — genre targeting in CTV will remain a signal that arrives often but says little.

via wurl.com (Original)

Filed under

Share this article:

More from Amara Osei

Amara Osei

Show full bio

News editor covering people, launches and day-to-day practice in marketing, advertising, marketing technology, media, SEO and search at Marketing Herald.

177 articles

More on the wire

« Previous article