How Nate Silver Could Help Digital Marketing
You know I am passionate about digital marketing. But you might not know that I am also a baseball blogger. I am cmathewson at the Minnesota Twins blog Twinkie Town, one of the SB Nation sites that recently went through an unfortunate redesign. I don’t write there very often anymore, in part because of the redesign. But at one time, I was one of its most prolific contributors, when the baseball world was going through a controversial culture clash between insular scouting paradigm to one based on math. In the center of that controversy was Nate Silver.
I have been a Nate Silver fan for almost 10 years, when he developed PECOTA, a system that uses statistical analysis to forecast the performance of baseball players based on their past performance. At the time, what passed for baseball analysis was performed by “baseball men,” scouts who had grown up around the game and learned its nuances. Most of their insights were based on what they saw with their own eyes and their gut feelings. Silver was one of the young Turks of a sabermetric revolution among baseball analysts, people who used math to analyze players and predict their future performance, often more accurately than the baseball men.
Eventually, sabermetriccs became an established practice in baseball. Every team uses methods developed by Silver and others at least as a check against the errors of their scouts. Most rely more heavily on math, using scouts to fill in the blanks. Billy Bean of Moneyball fame is one such baseball executive.
Apparently, Silver loves controversy. When sabermetrics became an established practice, he set out to do to political analysis what he helped do to baseball analysis, with his fivethirtyeight blog in The New York Times. Political analysis has long been ruled by the scouts of the politics world, aka pundits–highly educated men (mostly) who glance at polling data and form intuitive opinions about them. Until the 2012 elections, pundits ruled. But in this election, Silver ruled, correctly calling all 50 states a week before the election and getting very close on the popular vote. According to Silver, the electoral vote would be a landslide for the President. The pundits called it a “tossup” using such mixed metaphors as “razor tight.” More pundits predicted a Romney win than an Obama win.
Some say there was one true winner in the 2012 election and it was Silver. No pundit in the history of political commentary was as accurate as Silver was on that night. I don’t recall an election in which the old guard pundits were so far off, either. The contrast was stark. He was so good, he managed to make his brand of analysis an established practice in less than half the time it took him to do the same thing for baseball. In fact, Silver’s success in the 2008 election influenced enough people that his models were in wide use behind the scenes in the Obama campaign. Future campaigns will take notice, and ignore the kind of analysis Silver performs at their peril.
Since Silver likes controversy so much, perhaps his next challenge can be digital marketing. Digital marketing is in the throes of the same kind of conflict baseball and politics have gone through. Mad Men on one side, geeks on the other. The Mad Men flaunt their experience using their particular brand of creativity to develop and push content on unwilling masses, hoping for a small percentage of them to engage. Geeks work to learn the willing audience in detail and target them with content that helps them make smarter decisions. I fall firmly in the geek camp. But I struggle to convince the Mad Men. Enter Silver, who could cut through their BS and transform digital marketing.
If you’re interested in how, please read on.
Prioritizing content marketing efforts
One of my favorite parts of the fivethirtyeight site is the Return on Investment Index, a map that shows in which states candidates investments have the highest returns. There was a marked difference in spending by the candidates. The President spent the lions share of his war chest in the states Silver identified, such as Ohio and Virginia. Meanwhile, the week before the election, the Romney campaign and the GOP Super PACs flooded the airwaves in states like Pennsylvania and Minnesota, in which they had little chance of winning, wasting millions of ad dollars on a losing effort.
As I’ve said numerous times, I sit on the Google Tech Council, a consortium of representatives from the largest B2B technology companies. All the people around the table at our recent meeting cited staffing and budgets as the two biggest challenges they face. Many of their companies still spend more money interrupting their clients and prospects with TV ads than they spend on intercepting their prospects information tasks with relevant content. Suffice it to say that every dollar spent in digital needs a clear understanding of the return on investment.
Digital marketers can’t afford the kind of mistake that plagued the Romney campaign. They need the best data available to discover where and how to spend their limited budgets and expend their stretched staffs. The best data digital marketers have to learn how to inspire action from their audiences comes from their search queries. Properly analyzed, data from search engines drives investment, showing you which queries have the highest volume, the lowest competition and the strongest relevance for your target audience. It seeds better social listening (garbage in, garbage out). It helps you predict the size and scale of your marketing efforts. It can even drive product strategies, branding and naming efforts.
Silver’s history is to model available data to make better predictions. In the case of baseball, it’s metrics like on base plus slugging (OPS), fielding independent pitching (FIP), and wins above replacement (WAR). In the case of politics, it’s statistically filtered polling data. If Silver did digital marketing, he would use keyword data extracted from customers’ search and social behavior. Each of these data sources is flawed,which has led their detractors (scouts, pundits, Mad Men) to dismiss them. Silver has convinced even the detractors that the first two can yield accurate results with the right math. The third data source needs similar treatment.
I don’t really expect Silver to take up the challenge. So I will ask, “What would Nate do?” and take up the challenge myself.