Dialling up CTV Performance with Audience Plus Context

In this contributed article written exclusively for ExchangeWire, Bill Swanson, EMEA strategy lead, IRIS.TV, discusses how contextual video data and the unification of linear and CTV can fuel omnichannel growth in the privacy-first era.

The mass consumer migration from linear TV to CTV has inspired advertisers to evolve their media buying strategies from content-first to audience-first, mirroring their approach to web ads. 

Bill Swanson

Bill Swanson, EMEA strategy lead, IRIS.TV

But these two strategies shouldn’t be mutually exclusive: CTV advertising would actually benefit from a hybrid strategy combining both audience and content targeting, enabling more accurate campaigns with higher performance. 

So how can the industry fuse these methods to achieve the best of both worlds?

Leverage the power of data at the video level 

Linear TV advertising can reach audiences at vast scale in a brand-safe environment, such as alongside the World Cup Final or primetime news. Despite its impact, however, linear TV’s wide net generates a high degree of wastage. For example, ads for pharmaceutical products aren’t relevant to most viewers of Grey’s Anatomy. CTV fills in the gaps by providing data that allows inventory buyers to filter relevant audiences, thereby targeting more refined demographic segments. But in order to maximise its potential, CTV must also offer data at the level of individual videos.

Previous efforts to understand video content involved analysing page URLs, but this can’t be applied to CTV since there is no web page URL to analyse. Even when analysing webpages previously, research across tens of billions of impressions found that programmatic ad impressions based on the webpage header failed to match the video content 60-70% of the time. How many times have you read an article and found that the video is only partially relevant to the actual article you have been reading?

Rather than use this outdated approach, the industry can overcome this lack of contextual transparency for CTV by working with trusted partners to extract video-level data. This includes the description text shown when browsing shows, such as on ITV Hub or Amazon Prime Video, and also video transcripts, which provide more granular detail about the content. Content data, which was originally siloed in proprietary platforms and databases, then needs to be unified by a connector, which provides common access to publisher content management systems, such as those managed by Tastemade or Verizon. 

Publishers should avoid a repeat of the situation with programmatic display advertising on the web, where data would be collected from a webpage, sent to a DSP, used to train algorithms analysing performance, then sent to the bid stream where the source lost ownership of it. What’s needed is a pipeline that only shares pockets of anonymised content data with the various parties, protecting the publisher’s control of its data and enabling advertisers to target and buy suitable video inventory based on brand-safety analysis. All of this can be actioned without touching personal data/IDs or cookies, just as it is used as a buying metric in linear TV--this is actually evolving linear buying techniques and enhancing them by utilising programmatic pipes for more efficient and effective execution.

Use context-based targeting

Advertisers adopting CTV into their marketing mix need to consider adding context-based advertising to their existing audience-based approach. Once video-level data has been extracted, advertisers can purchase inventory based on show content. This enables them to reach their target audiences while also improving the user experience--but there are two fundamental linchpins needed to achieve this.

Firstly, neutral video data platforms are required for publishers and distributors to add their video data to whilst maintaining control and preserving user privacy. These platforms must allow the passage of data using multi-faceted APIs, enabling contextual information such as metadata, as well as video and audio elements, to be transparently shared with SSPs, DSPs, ad servers, verification providers, and content assessment tools. 

Secondly, data must be amenable to analytics tools that can mine for deeper context. By implementing granular scanning of audio and motion data, these tools can simultaneously evaluate multiple videos, enabling fine-grained placements while filtering out unwanted inventory. Using a system trusted by both publishers and buyers, contextual and user-friendly targeting at the video level opens up new possibilities for greater relevance, better user experience, and higher ROAS.

Unify linear and digital teams

Linear TV teams, with a content-first mindset, and digital video teams, with an audience-first approach, have historically operated separately within the agency holding company model. Meanwhile, CTV has existed between these two realms. As marketers shift their focus from linear TV to programmatic auctions in CTV, they face a completely different supply landscape where content is fragmented across many different publishers. The risk is that this fuels confusion around CTV, and marketers are discouraged from investing in it long-term.

The future lies in the convergence of linear TV and CTV/Digital teams, working together transparently with different inventory, while sharing the same campaign strategy. Omnichannel viewing has recently become an important trend for brands, with publishers packaging streaming services with their linear TV offerings. In fact, NBCUniversal merged their linear and digital ad sales teams way back in 2015. Collaboration between linear TV and CTV means campaigns can be better planned from the beginning. Therefore, a unified media strategy across all video channels increasingly makes sense. 

The future of CTV

Any restructuring of ad strategy can seem daunting, but the reality is that CTV is here to stay (just consider your own changes of media consumption over recent years), and the market share of video streaming will only increase over time. Ensuring a winning combination of audience and context-based buying requires the presence of video-level data to analyse on-demand video inventory. Rich APIs and neutral networks provide the tools needed to target the granular context of this video data. And finally, breaking down data silos, and unifying linear TV and CTV ad teams, completes a holistic strategy across all video channels. By following these practices, the industry can successfully combine linear TV and CTV to dial up campaign performance in an omnichannel world.