Most consumers will not be aware this is happening - and may be creeped out if they are - but it appears that people have become inured to the tracking of their behavior. The question is whether there is a limit to human acceptance of these practices - and, if so, where it may be. JL
Stewart Rogers reports in Venture Beat, image by Pixabay:
AI is showing where to make changes to a campaign so it hits its exact target market — and when to pull the plug. The AI can determine whether the video is eliciting an emotional response during playback. AI is also helping brands determine where the high and low points of a video sit, which can be used to create powerful trailers and social ads. By better understanding which parts of a video effect which audiences and genders, brands can save time and resources.
Marketers have plenty of ways to measure video campaign success, but artificial intelligence is uncovering new methods for determining whether the dollars you’re spending are being applied optimally.
That’s what video insights company YouFirst is offering, and a recent study of one of its clients, spanning 13 video campaigns over two years, is revealing. Even with the latest and greatest analytics tools at the brand’s disposal, AI and emotion tracking are opening up new insights.
More importantly, AI is showing where to make changes to a campaign so it hits its exact target market — and when to pull the plug.
YouFirst works by allowing a focus group of video viewers access to the content through its player, which — with permission — monitors the facial expressions of the consumer via a webcam.
The AI and platform YouFirst developed can determine whether the video is eliciting an emotional response during playback, and it can understand the difference between six main emotions; anger, disgust, fear, happiness, sadness, and surprise.
In its investigation of the brand’s videos, YouFirst discovered something interesting and, ultimately, money-saving for the client organization.
Despite the brand targeting a mainly female audience, the data showed clearly that its initial campaigns were scoring much higher with a male audience. As all marketers know, after these insights are surfaced, they still need to be acted upon.
“I am a little afraid to [say] how much these results reveal,” YouFirst CEO Lujza Bubanova told me. “While ‘well put together’ questionnaires may lead to the satisfaction of all parties, emotions can point out some serious problems with concepts, targeting, or production. Who is going to be the one to [say] this can’t go out when that happens?”
And while the content being measured did hit a sweet spot during video production, it was the use of AI pre-launch that helped the customer make the right decisions. Here, for example, is a report produced by YouFirst that shows the emotional response, broken down by demographics, of an early video in the series.
“Traditional research showed that youngsters recognized the brand very well,” Bubanova said. “However, the emotions expressed are telling us it is boring to them. No ‘sad’ trigger is present in watching the video, and we look sad when we’re bored (the whole face goes ‘down’ — eyes, lip corners, eyebrows). Engagement for this group is also shallow, and since we measure engagement by taking into account head turns, we knew viewers were looking around and not concentrated on the video.”
Compare that to the most recent campaign — in this case, a gardening video. While the older female group responded most strongly to the content, it is evident that their interest and emotional response is nowhere near the level of the earlier campaign.
YouFirst’s AI is also helping brands determine where the high and low points of a video sit, which can be used to create powerful trailers and social ads. In the first instance, we can see that the video helps brand recognition by loading all of what YouFirst calls “power moments” toward the end of the video.
In the latter, we see that the brand message is associated with sadness. Depending on the vertical, that is likely to be the opposite of the desired response. Of course, many charity organizations rely on generating sadness to increase donations, but in this case, the goal is to sell more products. Sad people don’t buy.
What is unique about this study is the long-term nature of the data. These video campaigns were released across a two-year period, and the insights made available through AI and emotion detection have helped the brand save money in the production of new video content. By better understanding which parts of a video effect which audiences and genders, brands can save valuable time and resources from the outset.
The client’s strategy has been bold. Targeting one small demographic, especially when the delivery channel is so broad, is not for the faint-hearted.
“Especially with TV ads for mass products, the strategy to be liked only by one target group is quite daring,” Bubanova said. “In our study, we point out that the ads are striking to only a female audience of over 46 years old.”
Ultimately, AI is continuing to open up new ways for marketers to make decisions, but YouFirst sees this as just one industry that can take advantage of emotion tracking in the future.
“Emotion tracking will be part of any decision-making engine, whether it involves one’s social feed, whether it is in programming, whether in production, trend-setting, or discovery, down to even defining price sensitivity,” Bubanova said. “Even hedge funds use and will use more of it to decide where to invest.”
In the end, that’s the difference between smart, personalized marketing spend and spending to deliver advertising volume in the hope that someone listens.
“The TV industry is under a lot of pressure,” Bubanova said. “The world is saying that online advertising will push it out. I believe there is not much of a difference between the two channels, as the motivations behind why people watch their devices — either as the main screen in a living room or a small device in the hand — remains the same. YouFirst is bringing monetization opportunities closer to creatives by leveling up research methods, deriving trends directly from viewers’ facial expressions for the mass media industry to have a way to stay relevant.”
0 comments:
Post a Comment