Why Signal Quality Is Shaping the Future of Streaming Media Performance
Streaming has quickly become one of the most valuable environments in digital advertising, largely because of how people engage with it. Viewers arrive with intent, choose what they want to watch, and tend to stay with content for longer. That shift in behaviour has drawn budgets in the same direction, with more investment moving into video and connected TV where attention feels more deliberate and less fragmented: in May’s quarterly revenue report, Disney announced a significant 5% increase in advertising revenue, partly accounted for by higher impressions linked to streaming.
As growth in this sector has taken place, the volume of data used to plan and optimise campaigns has expanded alongside it. Advertisers now work with a wide mix of signals, including contextual cues, viewing behaviour, and various forms of audience modelling layered across different systems. While this should, in theory, create a clearer picture, it often introduces the opposite. Signals begin to overlap, and in some cases contradict one another, which makes it harder to understand what is actually shaping performance.
The distinction that starts to matter here is between signals that reflect something real and those that rely on assumption. When both are treated equally, clarity tends to fall away.
When more signals make decisions harder
The streaming ecosystem is built on several layers of interpretation, each adding its own perspective. Platforms capture viewing patterns, publishers provide metadata, and third-party tools apply classifications and audience segments. All of these contribute to how inventory is understood, although they do not always align in a consistent way.
Much of the data used in targeting remains indirect. Metadata can be incomplete or too broad to be useful in isolation, while audience segments are often constructed from limited inputs and then scaled across large groups. Content itself is frequently reduced to general categories, which can smooth over the detail that gives it meaning.
That lack of precision carries through into campaign delivery. Advertisers can struggle to reach audiences with confidence, while suitable inventory is filtered out before it has a chance to be considered. Optimisation becomes slower, because the underlying signals are not always reliable.
This is particularly noticeable in video. Despite the quality of streaming environments, over half of inventory never makes it into active buying. In many cases, the issue is not suitability but understanding. When signals feel uncertain, decisions tend to become more cautious, and valuable content is left unused.
Understanding intent through content and behaviour
There has been a gradual shift in how advertisers think about data in response to this. Rather than focusing on the volume of signals available, attention is moving toward those that reflect behaviour more directly.
In streaming, that often comes back to the content itself and the way audiences engage with it. A viewing choice carries a clear indication of intent, particularly when it is supported by how someone interacts with that content. Together, these signals offer a more grounded understanding of what matters to a viewer at a given moment.
This has implications for how campaigns are planned. Instead of relying heavily on inferred audience profiles, advertisers can align their activity with content that people are actively choosing to watch. That alignment tends to feel more relevant, which in turn supports stronger engagement and more consistent performance.
Premium media environments play an important role here, as they provide structure and consistency that help make these signals easier to interpret. The value lies in how clearly audience behaviour can be understood within those environments, rather than in scale alone.
Why streaming needs deeper contextual understanding
A large part of the challenge in streaming comes down to how content is interpreted. In many cases, classification still relies on metadata or transcripts, which only provide a partial view of what is actually happening within a video.
Important details can be lost in that process. Tone may shift throughout a piece of content, visual context can change meaning, and narrative framing often shapes how something is perceived. A scene involving a knife, for example, could appear in a cooking segment, a crime report, or a scripted drama, each carrying a very different implication depending on how it is presented. These elements are difficult to capture without analysing the content itself more closely.
Advances in AI are starting to change that. Video can now be analysed at a much deeper level, with systems able to process visual scenes, audio cues, and narrative context in combination. This creates a more accurate and complete understanding of what a piece of content contains.
With that level of clarity, decisions become more informed. Content that may previously have been excluded can be reassessed with greater confidence, while differences between similar topics become easier to recognise. A scene that shares keywords with sensitive content, for example, can be understood in its proper context.
For broadcasters and publishers, this opens up inventory that may have been overlooked. For advertisers, it provides clearer signals to support where and how campaigns are placed.
Moving toward clearer signals
As streaming continues to grow, expectations around performance are becoming more demanding. Advertisers are looking more closely at where campaigns appear and what they deliver, which places greater emphasis on the quality of the signals behind those decisions.
When signals reflect real behaviour and real context, planning becomes more intentional and optimisation more effective. There is also a knock-on effect on measurement, as clearer inputs lead to results that are easier to interpret and act on. That, in turn, allows budgets to be directed with greater confidence.
Streaming combines engaged audiences with high-quality content, which makes it one of the richest environments for signals. The challenge lies in making sure those signals are properly understood and applied in a way that reflects how people actually watch and engage.
As more budget moves into video and connected TV, this becomes increasingly important. The ability to recognise meaningful signals and separate them from background noise will play a central role in how campaigns are planned and how success is measured.
[Editor's note: This is a contributed article from Mantis. Streaming Media accepts vendor bylines based solely on their value to our readers.]
Related Articles
How are streaming content companies doing? Publishers, creators, and aggregators have traditionally made the majority of their revenue on subscriptions, so why does everyone want to talk about advertising? Ad-supported content is more affordable. Because most streaming consumers have maxed out their subscriptions, leaving the SVOD market saturated, advertising is more likely to be the dominant incremental revenue driver for streaming over the next several years, based on how media companies are talking about themselves and guiding investors.
26 Mar 2026