MediaCom Gauges The Emotional Impact Of Video Ads

MediaComIt’s one thing if a consumer views your video ad, but it’s another if they were receptive to your content.

In an effort to gauge that impact, WPP agency MediaCom forged a global partnership Tuesday with Realeyes, an emotion analytics platform that uses opt-in facial-coding technology to gauge user response to online video ads.

The goal: to introduce personalized creative and test consumer responses to MediaCom-created content, said Palle Finderup Diederichsen, head of EMEA at MediaCom Beyond Advertising, an arm within MediaCom that focuses on new brand storytelling and content opportunities.

It’s all in service of an emerging discipline around “emotion-based” measurement to determine ad effectiveness. Apple, for instance, intends to acquire Emotient, artificial intelligence technology that analyzes facial expressions to determine the emotional response to ads. And platforms like Sticky  heat map high-traffic publisher sites and gauge insights around attention metrics.

Realeyes’ technology, which was built by Oxford University researchers, will be used within MediaCom Beyond Advertising group’s content hub. Globally, that division has teams in 43 markets with a total headcount of more than 700.

Agencies and advertisers are interested in these new measurement techniques because, as Diederichsen pointed out, creating compelling stories in a digital environment is very different from doing so on television. 

“This basically means they need more content that is better or at least [more] tailored, probably at the same budget as before,” he said.

MediaCom Beyond Advertising is specifically focused on understanding the connection between brands, consumers and platforms, and tests new tools to aid in content creation. It also collaborates with vertical-specific divisions within MediaCom, such as MediaCom Sport.

MediaCom is already testing Realeyes’ tool for clients like P&G, Shell, SONY Mobile and Coca-Cola. Rather than measure campaign effectiveness after the fact, the agency wants to evaluate content in real time. That way, clients can either put more resources into or pull back on certain investments depending on whether a piece of content is resonating or falling flat.

Realeyes runs two forms of measurement: “live testing” campaigns are priced per view and prompt users who have opted in to “share” how they feel (e.g., happy, sad, excited) after clicking on a video in real time.

In “sample testing,” advertisers can select sample audiences to run pre-launch tests and are priced per sample. It offers emotion-based measurement APIs via key video players on YouTube, Brightcove and JW Player.

Emotional engagement is a relatively nascent KPI, but “has also been proven to correlate with sales,” said Diederichsen, “so that is an obvious goal for clients such as fast-moving consumer goods” companies.

Although Realeyes is rooted in desktop video, more consumers are viewing mobile video, which could introduce some limitations with the tech especially around socially distributed video.

“Mobile is already integrated [and we can] embed videos in live environments such as Facebook,” Diederichsen noted. However, he added as an example, “we always need permission to test [using this methodology] and this is unlikely to happen at scale with people who are on their daily commute on the train.”

 

Correction: Realeyes develops facial-coding technology, not eye-tracking technology, as the article earlier stated.

2 Comments

  1. I don't understand the hype around emotional measures. We really don't need more vanity metrics related to engagement in times where we can easily connect offline behavior to digital campaigns. If someone is happy, sad or angry - why should I care if they don't buy my product? Only the connection between sales and behavior matters.

    Reply
    • Hi JayJay - emotions aren't simply vanity metrics, they very much do drive customer behaviour. We can link emotional engagement to customer activity, from sharing to sales.

      Reply

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>