Turning personalisation into profit
FEED explores personalisation’s impact on engagement and the challenges broadcasters face in adopting addressable advertising at scale
|31.10.2025
5 min read
Addressable advertising is revolutionising broadcast by delivering personalised ads to segmented audiences in real time. This shift is not only enhancing viewer relevance and engagement, but also offering advertisers improved return on investment. However, the adoption of addressable advertising at scale presents a complex set of technical and operational challenges. We explore the technologies enabling this shift, its impact on viewer engagement and the challenges broadcasters face in adopting data-driven, targeted ad solutions at scale.
According to Chris Singleton, director of innovation, strategy and corporate development at Hoppr, addressable advertising only works when logic and data sit together, inside the operator’s environment and in real time. The industry’s challenge isn’t just inserting ads; it’s keeping the data, the decisioning and the delivery connected without breaking consent. “We’ve taken an approach that’s built on a simple but powerful stack,” he begins. “Darwin decides who to reach, Warren decides what to show and SSAI (server-side ad insertion) is how it’s shown.”
So, what does that look like in practice? “Darwin is the audience intelligence engine running inside the operator’s walled garden,” explains Singleton. “It uses deterministic, privacy-safe signals – such as what’s being watched, when and on which devices – to create real household profiles.”
Warren is the monetisation control layer that turns Darwin’s ‘who’ into a commercial decision: which ad to serve, in which slot and with what frequency rules. “SSAI provides the seamless server-side stitching that ensures the viewer sees one clean, uninterrupted programme stream. When you keep it all inside the operator’s ecosystem, you get relevance without data leakage. That’s the difference between just another ad server and an addressable TV model that actually works.”
James Varndell, senior director of product management at Bitmovin, argues that contextual targeting is increasingly being used to enhance addressable advertising, often alongside first-party data. “This methodology allows video providers to deliver ads that are relevant and targeted, without breaching user privacy regulations,” he explains. “AI-driven technology is being used to analyse video and audio content to provide personalised ads for viewers that are based on the content they are watching at that precise moment.”
Bitmovin’s AI Scene Analysis embeds multimodal AI directly into the VOD encoding workflow, combining large language models (LLMs) and computer vision to analyse every scene for mood, setting, characters and narrative flow. “The system automatically generates structured metadata that can be used to deliver contextually relevant ads for increased ad relevancy and engagement.”
Viewer engagement in addressable advertising
Viewers are naturally going to be more engaged with ads if they are of interest and relevant to them personally – this is what addressable advertising helps to achieve by delivering precisely the right ads to the viewer, on an individual level.
“That said, it’s worth remembering that ads are more effective if they’re delivered at the point when the viewer is most receptive,” says Varndell. “AI-based predictive analytic tools can be used to predict user engagement and conversion rates at different points of the video, so ads can be placed when the viewer is most likely to convert. At Bitmovin, we combine this approach with AI-based contextual advertising so that ads are both relevant and placed at the optimal point.”
For Singleton, the ways in which addressable advertising have improved viewer engagement is simple. “Relevance reduces friction,” he says. “When an ad aligns with a household’s actual interests, it feels useful, not like you’re being forced to pay an ‘ad tax’ before getting back to your show.”
Take a household that streams travel content – they might see an airline ad while watching the news. The ad follows the audience, not just the channel, so it feels natural. “You’re shifting the model from ‘interrupt and hope’ to ‘assist and measure,’” says Singleton. “The result is fewer wasted impressions for advertisers, better brand recall for viewers and a premium, uncluttered environment that broadcasters can defend and monetise effectively. It’s a genuine win-win.”
One of the important steps in segmenting audiences for targeted campaigns is working out viewing behaviours, and Singleton highlights that the big shift is ultimately from guessing to knowing. “Panels and proxies are out and deterministic, first-party operator signals are in,” he says. “That’s where addressable advertising really comes into its own. You’re now building segments from household reality.”
It starts with real viewing behaviours, the device graph in the home and the contextual cues such as time of day and genre. “From there, you can move beyond crude demographics and apply psychographic frameworks such as ‘Value Seekers’ or ‘Progressive Families,’” adds Singleton. “These clusters are far more useful for a brand’s brief, all while keeping privacy intact. Crucially, those same segments need to be activated consistently across the entire TV and digital ecosystem. Because it’s the audience that travel, not the channel.”
Server-guided ad insertion
There are, however, many obstacles broadcasters face in adopting data-driven, targeted ad solutions at scale. Delivering seamless and targeted ads at scale is a complex and challenging task. “If providers get it wrong and ads aren’t well-timed or playback is clunky, or if ads feel irrelevant, viewers get frustrated and are quick to abandon content,” says Varndell. “The recently published Bitmovin Video Developer report shows that ad insertion is one of the top three biggest challenges for video services.”
From a technical viewpoint, service providers face some major challenges: for example, ad blockers can prevent ads from being delivered when using client-side ad insertion (CSAI), and playback on specific devices can be challenging with server-side ad insertion (SSAI). “Errors can also occur at the transition between ads and the main content in a stream, particularly if the main content is encrypted and ads are not,” continues Varndell. “To overcome these issues, many service providers are turning to SGAI (server-guided ad insertion), which combines elements of both CSAI and SSAI.”
Additionally, to get the best return on investment in advertising, broadcasters require data to measure ad performance. “This includes viewership data such as the number of ad plays, ad abandonment rates and plays per ad quartile,” explains Varndell. “Alongside this sits quality of experience data such as ad errors, start-up times and rebuffering.”
Obtaining real-time observability data across advertising can be challenging, and for this reason broadcasters are now looking at solutions to collect this data from the video player directly. Varndell notes that this type of observability allows broadcasters to alter their ad targeting in real time.
Singleton suggests that the main challenges are 20% technology-based and 80% culture-based. “The tech debt is real. Most broadcasters still have siloed linear and streaming ad stacks, which makes unified measurement a nightmare. Our approach is to integrate with the infrastructure that’s already there, so you don’t need a costly ‘rip and replace’ project just to get started. But the cultural shift is the harder part, it always is. For 50 years, broadcasters have successfully sold time in a programme.”
Now, addressable advertising is asking those same broadcasters to sell an audience across moments. “That rewires the entire commercial model – from packaging and pricing to sale incentives,” says Singleton. “It’s everything and it’s a huge change, but the upside is real: advertisers will always pay a premium for verified, defined audiences and it unlocks yield face beyond the traditional schedule.”
The point of all this isn’t just to make the legacy TV ad model 10% more efficient. It’s about fundamentally broadening who gets to harness the power of television.
“When addressable advertising is measurable, brand-safe and viewer-friendly, it opens that door to the long tail of advertisers – the local car dealers, the regional travel companies and the niche e-commerce brands that could never previously justify a TV campaign,” Singleton concludes. “That’s new money for broadcasters and better outcomes for advertisers. It creates a Total TV ecosystem built on consent and service, not interruption. And that’s the real prize.”
Check out our November Signal here!
Share this story
