Streaming's Metadata Conundrum
The results are in: the streaming industry is not happy with the current metadata situation. That's a key takeaway from the recent Amagi x FASTMaster Intelligence industry survey around metadata. I wrote a detailed breakdown of the results and what they mean in the latest Amagi AirTime report.
I also put together a handy, interactive, overview of the results. It is important to note that these are directional, seeing it as qualitative research. But the results do speak to the frustrations that many in the industry have regarding metadata, as well as what it expects AI's role to be going forward. You can find that below. I hope it proves interesting.
Special study · Metadata in media operations
The Metadata Reckoning
What broken metadata is quietly costing the streaming business — and why the industry agrees on the pain but not the fix.
of senior media operators say poor metadata is costing them real money — lost ad revenue, weaker discovery, content pushed down the rail.
● 0% disagreed — the survey's one near-unanimous signal01 · The cost isn't the data, it's the delivery
The format tax
The single most-agreed statement in the survey isn't about metadata's cost — it's about reformatting it for every platform's bespoke spec. Only 39% think metadata itself costs too much. The pain is the toll of delivering it differently, everywhere. Click any bar for the full breakdown.
02 · Caught in the middle
The two-front squeeze
Distributors absorb mess from both directions: incomplete metadata arriving from content owners upstream, and a different reformatting demand from every platform downstream. 64% are hit on both sides at the same time.
03 · Consensus on the problem, not the budget
The action gap
Everyone agrees poor metadata is bleeding money. Barely half plan to spend more on it — and roughly a third have no clear investment plan at all. That gap is the opportunity.
Where the money is going
Among those planning to invest, AI tooling leads.
Who we heard from
A rare clean spread across every link in the chain.
04 · The bet on the machine
The AI bet — and the irony
A majority already trust AI to generate metadata with little more than a spot-check, and two-thirds expect it to run the job with minimal human oversight within three years. The irony: they trust the machine to write metadata faster than they trust their own partners to deliver it clean — which keeps getting worse.
05 · The ground is moving
Today → 2029
Operators see the discovery floor shifting under them — which doesn't make metadata less important. It makes it the fuel the recommendation and agent layer runs on. Bad metadata won't just look wrong in a grid; it'll be invisible to the systems doing the recommending.
06 · The thesis, in one line
Metadata is still treated like back-end hygiene, when in reality it should function as a front-end monetization asset - because monetization starts the moment a viewer considers what to watch, and metadata determines what gets surfaced, selected, and ultimately monetized.
In their words
The open-ended answers, lightly cleaned and anonymized. They all circle the same drain: inconsistency.
Metadata, like any data, is useful if it's helping viewers decide. It isn't technical specs that drive users - it's the quality of metadata that can be consumed by agents, human and AI.
As a content aggregator, the biggest issue is the lack of consistency in how content partners define metadata fields - e.g. how they classify their content by genre.
Having to organize and standardize metadata from multiple content owners, as the channel owner.
Mapping different internal and external systems' data - being constrained operationally while trying to innovate for commercialisation, personalisation and CX.
Content partners delivering inconsistent or incomplete metadata.
Not refreshed enough. And [a leading metadata provider] is poor.
All the variations needed across platforms.
The read
Six takeaways
The format tax, not metadata itself, is the pain.
86% call reformatting per platform their biggest drag; only 39% think metadata costs too much. It's the delivery, not the data.
Distributors are in a two-front squeeze.
71% say upstream inputs are worsening while 86% reformat downstream. 64% feel both at once.
Everyone agrees it's bleeding money. Barely half are acting.
86% say poor metadata costs real money; 50% plan to spend more. The budget hasn't caught up to the consensus.
The trust leap to AI has already happened.
57% trust AI to generate metadata now; 68% expect it to run unsupervised by 2029 — even as human inputs degrade.
The discovery floor is moving — so metadata matters more, not less.
Metadata wins discovery today (61%) but 68% expect AI, social and feeds to overtake it. Metadata becomes the fuel that layer runs on.
Standard-setting power is sliding to the platforms.
68% expect platforms, not content owners, to dictate metadata standards within three years. Fix your house now, or inherit someone else's spec.
FastMaster Intelligence × Amagi
We built the interactive read.
Amagi wrote the report.
This portal is the headline. The full Amagi Airtime Report goes deeper — methodology, global FAST viewing & ad-impression data, and the operational playbook for fixing metadata before the platforms fix it for you.
MethodA pulse survey of 28 senior practitioners across the FAST & streaming value chain — content owners, channel operators, broadcasters, platforms, agencies and media-tech vendors — fielded June 2026. 64% North America. At n=28 this is a directional read, not a statistical verdict; two items (poor metadata costs money; inputs are getting worse) drew zero disagreement.
AboutFastMaster Intelligence turns industry data into editorial. This study was produced with Amagi for the Amagi Airtime Report. Figures are computed directly from the survey responses. Quotes are verbatim, lightly cleaned and anonymized.
