Recent investigations by marketing researchers into what connects best with consumers in the digital space reveal interesting, even surprising, best practices for beefing up targeting on Facebook and creating a highly tweetable Super Bowl ad.

Digital marketers have a love-hate relationship with data. On the one hand, it's exciting having access to so many factoids about how consumers engage with content because this data holds the potential to make marketers better at their jobs. On the other, trying to make sense of all this information and elicit meaningful insights from it can send practitioners' blood pressure rising.

Luckily, data scientists are around to zealously mine the data and unearth nuggets of insight with the potential to help marketers improve results. Below are synopses of two studies that were presented at the ARF Audience Insights conference in Jersey City, NJ, earlier in the summer.

AIG optimizes Facebook targeting for greater granularity

Insurance company AIG's Rapid Learning Lab tries to maximize profitable growth for the company's marketing efforts through advanced statistical analytics, machine learning and big data. In a recent test, it was able to drive 160x the return on ad spend by beefing up targeting on Facebook.

"Facebook is different in that it doesn't have the same granular data that is available in other media channels so we can't push it quite as far as we can with others," said Newcombe Clark, global director of AIG's Rapid Learning Lab, during a session in Jersey City about the research. "But it has unique data and a lot of levers that can be pulled to improve performance. It takes some work but the results are well worth it."

Using basic ads with stock images in both the desktop and mobile newsfeeds as well as off-the-shelf targeting options, the lab launched a direct response campaign for supplemental travel insurance for cruises. The effort leveraged retargeting and lookalike audiences, a unique offering from Facebook, choosing the top five segments out of 100 audience groups, each one slightly less like visitors to the brand's sites.

In order to measure sales lift and profitability, AIG partnered with Doctors Without Borders to create a holdout group by having the organization serve ads about its services to a portion of AIG's audience.

While the initial effort performed reasonably well overall, the results were mostly driven by the top lookalike audience while several very large audience groups flopped. AIG retooled its strategy to see if it could improve results from the underperforming groups rather than simply shutting them down.

The data showed some clear opportunities to drive efficiencies, such as shutting off Android, where 26% of impressions were being served but conversions were low. Overall, 96.4% of conversions came from desktop.

There were also indications some underperforming consumer segments could be shut off with the money reinvested in the strongest segments: Baby Boomer moms, outdoor enthusiasts and an income category above middle class but not the top level.

Finally, AIG also shut off ads to consumers below 34-years-old and above 65.

These moves delivered savings that were reinvested in Facebook offerings like optimizing along the sales funnel, targeting based on types of cruises and multivariate testing of different ad copy and images. The winning ad displayed a negative image showing a flight cancellation alert and copy urging travelers to protect their family vacation. It was 166x more likely to convert than the bottom performer on desktop.

Finally, AIG was also able to improve results by cutting back on retargeting, which had started to have a negative effect once a saturation point was reached.  

What makes a Super Bowl ad tweetable?

Pepsi's humorous "Puppy Monkey Baby" spot for Mountain Dew was the most tweeted ad during the 2016 Super Bowl game. "Why?" is probably a question racing through the minds of many marketers. Luckily for them, a unique collaboration between Nielsen's TV ratings, social and neuroscience divisions, as well as Twitter and Pepsi, set out to answer that exact question.

The Super Bowl is one of the biggest TV events of the year, which is why brands spent an average of $30 million on a 30-second spot in 2016, according to Nielsen. Ads that resonate with viewers can also see their message amplified across social media. According to Nielsen, viewers sent a total of 17 million tweets during the game. In a sign that they weren't just paying attention to the game, 27% of those tweets were about ads.

"Big moments in the game create lots of excitement," said Sean Casey, president of Nielsen Social, during a presentation in Jersey City about the research. "On Twitter, this is a lot more severe. We saw that when there is a great ad, there are peaks in activity as well as when there's a great play and during the halftime show."

In a ranking of the game's 30-second spots by how many original tweets they were mentioned in, the top six ads averaged over 70,000 original tweets while the bottom ads averaged 3,400.  Retweets were left out as were ads for TV shows or movies. This left a total of 14 ads with no one category dominating, suggesting the positive response was driven by creative.

For the neuroscience portion of the study, the ads were viewed by test subjects in a lab months after the game to measure brain activity and visible responses. The results showed a big difference in how viewers reacted to the top ads compared to the bottom ones. For brand appeal, there was a 12% difference between top and bottom ads, for ad memorability 20% and for brand memorability 26%.

When reviewing facial responses to the ads, viewers had a positive response to the top ads 18% of the time — a high number even for humorous ads, according to the researchers — versus 6% for the bottom ads.

In terms of memory activation, the strength of the message did a better job at creating a memorable ad than the length of the branding.

So what does the research say about Pepsi's ridiculous Mountain Dew ad? Well, it over-indexed on all measurements. It was memorable and elicited a positive reaction. The ad's unusualness also produced a negative reaction, but not enough to detract from the smiles it ultimately delivered. The spot also showed more of what's called "wear in," meaning viewers have an intuitive sense that it will get better with subsequent views.  

The takeaway for marketers? It's important to test for emotional response before spending millions of dollars on a Super Bowl ad.