Can NBCU Really Measure Ads’ Emotional Impact?

NBCUniversal’s ambitions to reinvent measurement have now extended to embrace emotion and ad quality, but can some factors be too abstract to quantify?

What does a media company do after leading a charge to supplant Nielsen as the TV industry’s longstanding measurement currency?

For NBCUniversal, the next frontier is widening the measurable variables in an ad campaign. Kelly Abcarian, EVP, measurement and impact, advertising and partnerships, framed out the company’s ambitions in a blog post last week, writing: “We are exploring how we can value ad quality and embrace accountability across the board and connect the emotions ads create with the content to the impact it generates.”

The idea is being able to widen accountability for ads’ impact — namely, the efficacy of the creative in resonating with viewers and the ability of publishers to maximize ads’ reach on their channels. Agencies and publishers share the responsibility for impact, and the data is there to show it.

“As we go on this journey to impact, we wanted to ensure that we could start to bring together how we can protect our position as a premium publisher but really move toward making more complete data-informed outcome guarantees by contextualizing the end market performance of the creative decisions,” Abcarian says, her penchant for abstraction likely suited for the task in front of her.

After all, how does one measure some increasingly abstract variables in order to pin down the accountability NBCU is pursuing? How to quantify emotion with any useful accuracy, let alone scaffold that data out to connect to a business outcome?

NBCU initiated a wide request for proposals to kickstart that process, and Abcarian’s post singled out a number of companies whose research points to promising insights, among them Dumbstruck, System 1, Emoto.AI, Kantar and Dynata. Researchers there are using eye tracking, facial coding, neuron testing and even galvanic skin response (changes in sweat gland activity that reflect emotional arousal) towards emotional measurement.


“There is a whole arsenal of capabilities that we can unlock to try to better understand the emotional response that sits behind how consumers react to the creative that we then believe could have a correlation to better understanding the impact,” Abcarian says.

She says NBCU sees partnering with multiple external measurement partners on this front “so that we won’t become reliant on a single methodology.”

Initial discussions with these potential partners have kicked off a proof of concept, she says, “so that we can build that as a core of a creative measurement engine.”

Abcarian says the study of emotion has come a long way, yet concedes “we are early on this journey.” One major unknown in that journey is the scalability of this kind of measurement, and she hopes to convince others in the industry to collaborate on finding that answer.

“I think we will learn that … over the course of working with these measurement companies,” she says. “We are optimistic that we can create and feed in the metadata and results in which to create a true scalable testing that could be accessible to our entire roster of clients.”

So how might this data be applied to more efficacious ad placements by publishers? Abcarian says one use may be for the data to reveal “emotional pairings” that complement ads evoking specific emotions with ad breaks adjacent to scenes of similar emotions in a given program, thus holding the viewer by emotional engagement.

“That is definitely one application of many — how we can take the link between pre-market creative performance diagnostics and then bring those into market measurement,” she says.

As to when viewers might be seeing these calibrations playing out on their own screens, Abcarian says to stay tuned.

“As we move ahead, we will look to share results as we start to get all of that testing across a number of partners underway, some of which we have kicked off and some of which we anticipate kicking off in the coming quarters,” she says.

Given the goals for this kind of measurement, it’s only prudent to leave the implementation timeline a little abstract, after all.

Comments (0)

Leave a Reply