Connections

By Thomas Siegman, RSG Media Systems –

Data analytics is upending the media industry, helping companies buy, schedule, promote, and distribute content better, and to boost revenues from advertising, subscriptions, and sales. In 2012, the Obama campaign was significantly more effective at better targeting its viewers using a variety of innovative behavioral analytics techniques. By identifying the potential likely voters, they found that they could reach them through highly targeted, lower cost, late-night ads on ESPN and Nick at Nite, rather than traditional, more expensive prime-time networks spots.

In the four years since, the field has advanced much further.

Examples abound:

BSkyB doubles the likelihood of a viewer purchasing an advertised product by using credit-card data to identify target viewers and using the set-top box to insert specific ads; Netflix lulls viewers into binge watching by precisely analyzing the exact pause between episodes (if pause is too short or too long and viewers exit; time it just right and they watch for hours).; MVPDs price VOD offerings dynamically, altering how much they charge territory by territory, customer by customer.; Target uses purchase behavior to identify life instances, such as pregnancy, relocation, marriage, and divorce, when consumers’ behaviors are most susceptible to change.

In the digital world, where viewers spend 60 percent of their online time on a mobile handset, personalization is a given. Sites boost their CPMs by tracking viewers’ frequent shopper cards to profile users that respond to specific ads.

And yet, the overall industry lags. By our own estimates, companies conservatively leave 10 percent of earnings on the table. In a $2 trillion per year industry, that tallies up to $200 billion (more than twice the value of the entire Walt Disney Company) in lost revenues every year.

“For an industry that appears to be technologically sophisticated, their understanding of information technology is a bit startling” notes big data expert Jeff Needham, who regularly works with intelligence agencies on threat analytics and governance. He points out that the moment we digitized content, it became big data, ripe for analysis. “Content is still critical, but there no longer is a line between production, delivery, and sentiment analysis – it all happens on the same computing plant with many overlapping tools.”

The good news is that this is all changing. Although RSG Media has been in the data business for 30-plus years (25 in the M&E space), we have suddenly seen a massive uptick in media companies’ desire for data insights.

Three major trends are fueling this new interest:

Competition. 2015 brought viewers over 400 scripted shows on television, almost twice as many as in 2009. Pundits refer to this as “peak television” because this volume seems unsustainable economically. Everyone involved in producing and selling content is looking for an edge to remain relevant, attract views, and the dollars that go with them.

Investment Mindset. Anyone who invests looks to earn an ROI. Now networks are taking this same approach. Rather than merely generating hits, savvy programmers view their entire content asset library as an investment portfolio.

Having invested billions on content, they realize that earning an extra 2 percent in ads, VOD fees, or in other areas will contribute hundreds of millions of dollars to their bottom line.

Technology. New platforms allow us to interrelate huge amounts of seemingly unrelated data. In the past, we looked for everything to be in neatly structured tables. Now we can use the same techniques that the NSA uses to identify “threats actors” to identify “fans of actors” or other likely viewers, and how to persuade them.

A return to the art of selling

For most of human history, selling was a learned skill. Clerks spent long apprenticeships learning to size up a customer, identify their needs, their propensity to spend, and finally, to price goods for sale. Mass marketing is an invention of the 1900s; large department stores such as Macy’s, Wanamaker’s, and Boucicaut were able to thrive by undercutting rivals on price. Today’s marketers and media companies understand that price is no longer the predominant factor.

Data analytics gives this insight. Used properly, data can help media companies understand the viewer’s behavior better than the viewer herself. Rather than simply grouping viewers into demographic-based “look alike” models, we prefer “act alike”: what she watches; which ads resonate; how she buys; where she gets her information; what she tells her friends (and which of her friends listen).

For example, by combining viewership and revenue data from linear and digital platforms with data feeds from credit cards, Web and in-store purchases, social media, and more, some of the attributes that we can consider include:

• Affluence
• Age/life stage
• Brand loyalty/early adopter
• Demand
• Family makeup
• Fashion
• Financial strategy
• Gender expression
• Home type
• Influencers/followers
• Lifestyle & habits
• Location
• Purchasing propensities
• Social communities
• Vulnerability (finance, health, other)

Moreover, as the “Internet of Things” expands—the ability of everyday appliances such as our refrigerator, our car or our electric toothbrush to communicate on the Internet—we’ll have further, more detailed data that allows us to get inside viewers’ decision loop.

In a recent conversation with the CTO of one of the largest cable networks in the world, he described a future in which he will have up to 27 different feeds for the same network. That is, if two people tuned into the same TV network, they would see different things, perhaps even different shows; definitely different promos and commercials. All of this is based on data analytics and actalike modeling.

The challenge in creating these models is that it’s not enough to have data, or to bring it together into a large data lake. One needs to know what the data means, how to interrelate it, and how to discern actionable insights.

Data and the great unbundling

This data insight is critical because the competition is about to get even more intense. Already we are seeing the start of the great unbundling as networks that once swore they would only be available through an MSO now race to provide their signal “over-the-top” (OTT) direct to consumers. B2B businesses are diving headlong into the B2C space.

It does not take a crystal ball to see that, soon, programmers will sell shows a la carte. Yes, networks will resist it. In the end though, customers will balk at paying for an entire network when all they want is a single show. Networks will then brand and market their content just as any other consumer product, selling Game of Thrones the same way that Kellogg’s hawks its latest cereal.

The one fillip is that, more than selling content, networks sell viewers’ “eyeballs” to advertisers. Whoever is able to attract and package the best collection of eyeballs will be able to generate the highest revenues.

Earlier this year, we were able to demonstrate to a network that they could dramatically increase the efficacy of their promos by sniper-targeting their audience. We showed them how to focus exclusively on those whom we could convince to watch the show.

This meant ignoring both the people who could never be convinced and those who already were planning to watch the promoted show. The result was that they achieved higher conversion, a larger more engaged audience, with 30 percent fewer promos.

This is just the beginning. Retail pioneer John Wanamaker decried famously that half his advertising dollars were wasted; he just did not know which half. Now we are entering a “brave new world” with the ability to anticipate and create desire for content, products, and viewers.

The real question is how companies will now act to take advantage of this opportunity. Based on our work, we recommend six steps:

1. Appoint a data champion. Too often data sits across, or even outside the organization in different silos, managed by people who are reluctant to share. Companies need someone empowered to marshal all the data, wherever it may reside.

2. Understand the data. Hiring a group of Ph.D.s will not get you results unless they are also expert in the business of M&E.

3. Bring the data together. Modern data techniques allow the interrelation of wildly different and often unstructured data. The first step is to get all the data in one place.

4. Ask the right business questions. There is a fine line between correlation and causation. (The popularity of organic food has grown over the past five years; so have obesity levels, yet no one would argue that organic food causes obesity.) Knowing what questions to ask means that the answers will be meaningful.

5. Test for negative cases. Knowing what not to do is often as, or more important than, knowing what to do. Knowing not to advertise shampoo to bald men saves money.

6. Serve the customer. With great insight comes great responsibility. By utilizing data properly, one can help one’s customers have a deeper, richer, happier life.

Click here to translate this article
Click here to download the complete .PDF version of this article
Click here to download the entire Spring 2016 M&E Journal