Defying Gravity

Network television is at risk of getting caught in a vicious cycle.

As the audience fragments in a million different directions, smaller subsets of that audience see promos for new shows. Then, as new shows draw smaller crowds, even fewer viewers see promos for other programs.

The reach of television networks (the total number of viewers who watch for a minute or more once a day) is down a daunting 12 percent in one year. Yet a six percent larger audience has seen the promos for MTV’s Viacom networks—even though they’re using fewer spots. That is a stunning gap of 18 percentage points.

“When network reach is down, increasing your reach six percent is actually a huge accomplishment,” says Oktay Arifkhan, senior VP of analytics and measurement science for Viacom Media Networks.

Credit goes to the machines. Machine-learning algorithms are helping MTV and its siblings, Comedy Central, VH1, Spike, and others, show their promotional ads to the “right” viewers (the most impressionable and like-minded). They target new recruits with better aim and achieve broader exposure by mixing and matching the right networks and shows. Overall, they’re able to reach more viewers while using less ad inventory to do it, and they’re achieving some stunning increases in “conversion rates”—the portion of viewers who see a promo then tune in to watch the promoted show.

The Viacom networks have pulled off this death-defying stunt using Prophet 2.0, a “promo optimization platform” developed by RSG Media on behalf of MTV Media Planning 360, the multiplatform marketing and promotions department for MTV Networks. RSG Media is the sponsor of this series on Big Data Insights.

Prophet is the inspiration of Chris Visceglia, VP of media planning and scheduling strategy at MTV 360, who wanted to meld “the science and the art.” “The science is in how we deploy advanced algorithms and big data to drive our media planning,” he says. “The art is creating the best viewing experience possible and by always putting the viewer first in how we present creative.”

A study that Arifkhan unveiled at a Media Insights & Engagement Conference in Ft. Lauderdale this February revealed just how well Prophet does at placing promos and predicting audience patterns. Arifkhan says these results are real—this is not a test: MTV was able to reach 28.3 million people with promos on MTV for the 2015 MTV Video Music Awards, despite the fact that they used 17 percent fewer spots than they did a year ago.

By running VMA promos on sibling music channels MTV2, VH1, and Logo, as well as on stablemates Comedy Central, Spike, TV Land, and Nick@Nite, MTV expanded the reach by 40 percent, to 40 million viewers. The Prophet smart machine did so while using eight percent fewer total promo spots than in 2014.

In addition, by aiming VMA promos at viewers of shows on other networks that are simpatico with the MTV bent, the optimizer scored a much higher conversion rate than the year before: 24 percent of viewers ages 18 to 49 who saw the promos watched the VMAs in 2015, up from less than 20 percent of promo watchers in 2014.

Devoting five percent of spots to non- MTV networks resulted in expanding the reach of the promos only four percent beyond the MTV audience. But when one-quarter of the campaign was devoted to other networks, the reach soared by almost 30 percent. That translates into a seven-fold increase in reach for a five-fold increase in airtime.

A startling 47 percent of viewers who saw VMA promos across multiple networks went on to watch the awards show, as compared to just 20 percent of viewers who saw a promo only on MTV.

“That shows the value leverage of cross-channel promotion,” Arifkhan says. “When people see your message not just on your network but on multiple networks, it substantially increases conversion rates.” Since the 1950s it has been an unwritten law that a viewer has to see a promo three times to “convert” and tune in. “For millennials, if you increase the frequency to six to seven it’s optimal.”

The stats themselves are striking, but it’s even more remarkable that Prophet can know these things at all, that it can track viewing patterns and ratings, measure changes, and unearth strategies out of such a massive pile of Big Data. This machine-learning system is able to run more than 110,000 different simulations for scheduling programs and promotional spots, all within minutes, and pick the best solution. It can schedule many campaigns simultaneously and ensure the most important ones get priority. The platform then assesses the results, adjusting on the fly.

“Our machine-learning algorithm learns from past successes, but more importantly, past failures, and automatically adjusts to those success and failures when moving forward,” says John Curran, a business-knowledge specialist at RSG Media.

In the days before the Prophet platform, some Viacom nets would media plan using historical Nielsen data and research team guidance, along with scheduling intuition. This allowed for no opportunity to course correct in flight. The Prophet runs through ratings data 24/7, produces a trend report in two days, and adjusts accordingly.

It would take a staff of six full-time analysts to try to perform these same feats, Arifkhan says, and they couldn’t do it as well. Plus, they would have to take time out to sleep and eat and recharge, while the machines can run 24/7. And however smart the system is today, by general machine-learning standards it will be a thousand times smarter only a few years from now—and then smarter still by another thousand-fold increase a few years after that.

“This was my team’s concept and vision for a very long time, but we couldn’t have done it without RSG Media. They made it real,” MTV’s Chris Visceglia says. “As Viacom explores ways to measure how our content is consumed across different screens, this is a giant step in the right direction.” He adds: “The possibilities with this seem somewhat endless at the moment.”

Resources:

Big Data Projects on the Rise (But Data Use Could Be Better)
Big Data and the Future of Business
How the Super Bowl Uses Big Data to Change the Game
Big Data Is the New Ground Game: How Ted Cruz Won Iowa
Cloud, Mobile, and Big Data Analytics Drive Latest Rounds of Megadeals