Reelgood's AI Discoverability Study Reveals a Big Challenge For Streaming
Attention follows content. Money follows attention.
That logic has governed streaming economics since the first ad-supported tier launched. What's changed is where viewers go to find what to watch — and that shift is creating a quiet problem for FAST publishers and their ad partners.
AI assistants are now a primary discovery surface. And it turns out they're also unreliable. A recent analysis from Reelgood found ChatGPT and Claude answering basic "where to watch" queries accurately less than half the time. The errors aren't random — free and ad-supported services like Tubi, Pluto TV, and Kanopy are systematically omitted, even when they're valid options.
Findings:
100 titles tested: 50 movies, 50 TV shows
Test date: March 5, 2026
ChatGPT accuracy: 43.76%
Claude accuracy: 50.21%
Reelgood accuracy: 96.89%
The invisible inventory problem
FAST business models run on reach. If a title lives on an ad-supported platform but doesn't surface in AI-driven discovery, the impression opportunity disappears before a viewer ever opens the app. It's not a content problem. It's a distribution problem one layer up.
Reelgood CEO David Sanderson framed it simply:
"If you license something amazing and you shove it in the basement and don't tell anyone about it, how much good did that really do you?"
That was about on-platform merchandising. The same logic now applies to the AI layer above it.
Five primary patterns identified:
Outdated availability — models report titles as streaming on services they've already left
Hidden paywalls — add-on channels like Starz or Paramount+ inside Amazon are listed as included in your base subscription
Free services ignored — Tubi, Pluto TV, Hoopla, and Kanopy are consistently missing from results
Rent vs. stream confusion — titles only available to rent or buy are listed as subscription options
Wrong version, wrong platform — when multiple versions of a title exist, models mix up which one is where
The gap is solvable — but who closes it?
The data infrastructure to fix this exists. The harder question is who in the ecosystem has both the incentive and the leverage to act. Streaming platforms want discovery. Advertisers want reach. AI companies want utility. None of them are directly accountable when a FAST title goes unfound. Until that accountability lands somewhere, the gap stays open.
Watch the webinar replay with Reelgood CEO, David Sanderson here
Get the SOS. Brief
The sharpest streaming intelligence, delivered to your inbox.