New Call-to-action
This might be TMI, but I once had shrink who liked to say (repeatedly), “feelings aren’t fact.”

I hate to break it to her, but neuroscientists and academics beg to differ, arguing something more along the lines of “feelings make facts.”

Studies show that people rely on emotions to make most decisions — including buying decisions.

  • fMRI neuro-imagery shows that consumers primarily use emotions, rather than features and facts to make decisions.
  • Advertising research reveals that the emotional response to an ad has far greater influence on a consumer’s reported intent to buy than does the ad’s actual content.

This means that we feel our way to reason, or as Douglas Van Praet, author of Unconscious Branding puts it, “Emotions don’t hinder decisions. They constitute the foundation on which they’re made!” Many other experts agree, you can use your content to make people feel things and, in turn, encourage them to do things.

Exciting (and slightly scary) stuff, right?

That’s exactly what why we talked to Devra Prywes, VP of Marketing and Insight at Unruly, an ad tech company that uses data, tech and emotional targeting to get video ads seen by people who will like and engage with them most.

If you’re feeling intrigued yet slightly uncomfortable — then read on!

OZ: How did Unruly start?

Devra Prywes: We started with something called the “viral video chart. ” Our founders created this website and they would track online video and rank it by share. We track views, like lots of other companies, but we’ve always really focused on shares because shares are a measure of social buzz. It’s a measure of what people are actually talking about.

It’s a dirty little secret that a lot of people outside our industry aren’t aware of. Views are usually paid for. Views are usually representative of budget but shares aren’t.

OZ: Right, all those people who buy followers are dismayed to find out those followers don’t share.

DP: You can’t force someone to share an ad. I don’t have data on this but I’m convinced that no has ever gleefully waited for the next online video ad to interrupt their viewing experience and said “Wow, who am I going to share this with?”

But the thing is, that people DO share ads. And that’s what we’re focusing on. What are these metrics, what gets people to lean in, engage with the piece of content? They stop what they’re doing and share an ad with either their network or even more narrowly, with one or two people. That’s what Unruly’s predictive algorithm helps us with, it helps us predict the virality of a piece of content.

OZ: How can you predict virality if a piece isn’t done? Or what value does it have if you’re using it after a video is already made?

DP: We can use it early on in the production phase and let the creatives know, the brand know, what types of emotions are rising to the top. We also measure things like brand recall and purchase intent so you can look at it through the business metrics.

We can say, “This is really engaging people. This is rising to the top. Do more of this.” Or, “Your audience really is kind of tuning out over here. If you’re looking to edit something out, you could go lighter on this.” We deliver a 30-40 page insight report in the production phase.

If we get content that’s already done, even if no edits can be made,we can still find out what people like about it. That can help inform cut down, too. We can say “here are the peaks and valleys of engagement, you might want to use when you’re making your TV commercial.” Or, “people love this area, this might be a GIF that you want to launch at some point.” We can also see what devices people use to watch content, and what social platforms they like to use.

OZ: Which gets to what you guys really focus on quite a bit, distribution, right?

DP: We really like to hammer this home, from the big screen in our living room to the small screens in our pocket content is everywhere. If you build it, it doesn’t mean they will come, you have to make your content discoverable so audiences will engage with it.

OZ: So, um, how do you do that, exactly?

DP: So, some ads are really great, some ads aren’t. But there’s always an audience that will like an ad more than other people. We can help find those people that are more emotionally engaged. Use third party and first party data and kind of model out a segment and reach people like that at scale. It’s kind of like a target within your target audience.

That’s called Unruly Custom Audiences and it was kind of a gamble when we first launched it. But, knowing what we know about emotional science and what we’d learned about sharing, we thought it would work. And it did! We saw 16-20% uplift in video completion rates in a more emotionally engaged audience versus the brand’s target audience. For a while, we saw upward of triple percentage uplift in purchase intent and loyalty at the same time as we saw fewer negative comments.

OZ: Whoa, that’s crazy stuff. What emotions are you guys actually tracking and how did you decide those were the right ones?

DP: We track 18 psychological responses. The most important thing when it comes to video engagement and sharing is the intensity of psychological response. It’s broken into the emotional, cognitive, and primal areas of the brain.Screen Shot 2016-05-18 at 11.42.57 AM

What’s great is that we’ve seen this from our own data, but there are academics out there like Karen Nelson-Field and Jonah Berger who are huge in viral and all of us, pretty independently, have come up with very similar findings.

OZ: And how do you conduct the tests?

DP: The tests are done digitally. We recruit a nationally representative sample of about 500 people so we can get statistically significant batches of different demographics.

We do some facial coding, but that’s not enough. By tracking micro muscle movements, facial coding can test for surprise or anger, smiling, but there are a lot of things that we track through the survey, through the panel, that you might not have a face for. Things like inspiration, amazement, warmth, nostalgia, hilarity—you might not have a face for those—but those feelings will help drive sharing, so we like to offer them up to advertisers. We have a very cool dashboard that we can toggle on and off and just see how these different groups respond to the same piece of video content. Sometimes it’s really different. It’s crazy, totally different emotions will rise to the top from people watching the exact same video.

OZ: Really? Was there an incident that changed the way you think about the complexity of human emotion?

DP: We all know that that people approach situations differently, and they can have very different emotional reactions to the same thing, But, it’s really interesting to see the data behind it, to see it on a screen.

There was a Reebok ad where Reebok had partnered with one of these mud run companies, it was called “Freak Show/Be More Human.” It was these men and women who these incredible feats of fitness, like normal everyday people doing extraordinary things. There’s this one scene where a woman is running and she has a man across her shoulder, like a fireman carry.Screen Shot 2016-05-18 at 11.46.36 AMFor women, it registered as an empowerment moment, left them feeling inspired and proud, younger men, millenials, are left feeling nonplussed like, “yes, women do sports,” but older men are feeling confused. Scenes like that are fascinating to me.

OZ: Wow. That really a great reminder for writers too, you’re not writing for yourself. You’re writing for different groups of people who bringing their own baggage and history to everything they do and see. How do creatives react to getting feedback from an algorithm? In my experience, that can be tricky.

DP: I’ll admit, I was nervous about that at first. I was worried that they might feel like we would handcuff them or rein them in. But really, our reports have been received as this guideline and this guidepost. Here’s an example. We do this annual Super Bowl insight paper. We met with one of the automakers and all its agencies. The creative agency meeting was the most interesting meeting we had because for them, it seemed that this data could kind of be therapeutic.

First, we kind of started with some general insights, right? You hear the responses, you hit them with intensity. And they say, “Okay. You’re telling me that I need to hit these responses with intensity to drive engagement. You’re telling me hilarity can be really fickle because it’s subjective.” That is fickle, it’s just the toughest one. How many people are really that funny? Everyone thinks they are. Then they ask “Well, what do you do then when the when the client wants to go for humor but keeps cutting your best spot, your best part, what do you do then?”

OZ: Which always happens.

DP: Right! So many advertisers do go for funny but it’s really hard to do funny by committee. It gets really watered down. This is where data can really support creative teams because it takes the whining out of it. It’s not you going to the brand manager arguing “But this is the best part, this is funny!” because you’re taking data from 500 people, looking at the peaks and valleys of engagement and emotion over the runtime of the video and it’s really easy to see where it peaks.

The creative can then say, “great, we could cut this part, but, this is the most engaging piece of your ad. If we cut it, people will be less engaged with it and less likely to share it and give you earned media.” Suddenly it repositions the entire conversation and then no more whining.

OZ: Sounds like a sweet win. Does it go further than that? Are certain emotions correlated with conversion or purchase intent?

DP: Right now it doesn’t look like there’s a single emotion that is driving … like, “make it surprising and people will buy.” It doesn’t quite work that way. What we have been seeing is that it comes down to the intensity of emotion. That’s just what’s really most important.

There’s this best-selling book by Daniel Kahneman and it’s called “Thinking Fast and Slow,” and in it, he explains the two systems in the brain, the rational and the emotional system. It’s the emotions that make us want to do things. They’re the secret sauce to make people want to buy, want to do things, want to try new things. Then, our rational brain kicks in and dissuades us. We’re still looking at this but I would say some of the social motivation might come into play here more.

For example, if the ad is triggering social utility, it’s about a product where people are sharing it because they think other people will find it useful. If it’s about a product, usually it means people are understanding some of the product attributes. But in general, we haven’t been determined a specific single driver.

OZ: Well do let us know if you find the Holy Grail! Thank you so much for the interview and we can’t wait to hear more at the event next week!

New Call-to-action