Silverpop - Data-Driven Marketers Have a New Superhero: Nate Silver
It appears you are using an older version of your browser. This site was developed to be progressive and future-compatible. Please take a minute to upgrade your browser for an optimal experience.
Skip to content
  • Subscribe:

Data-Driven Marketers Have a New Superhero: Nate Silver

by: Dave Walters (@_DaveWalters)
09 November 2012

Regardless of whether your candidate ended up winning or losing on Election Day, we now know two things for sure: Math is a better outcome predictor than partisan opinion, and Nate Silver is the coolest data geek on the planet.

For those not familiar with him, the short version is that Silver is a mathematician-turned-baseball-stats-turned-political-predictor. He's formulated a predictive model on his 538 blog that combines political opinion polls and seeks to boil out all the partisan politics to create a mathematically sound model. Perhaps the best explanation I’ve seen of his work was from a recent Chicago Tribune article:

“Unlike traditional pollsters, who put questions to a field of voters, Silver incorporates the averages of several polls and weights them based on factors like the past accuracy of the polling firm, the number of likely voters on Election Day and the composition of each state's electorate. He then runs multiple simulations of the results, which results in his probability forecast.”

Oh, and by the way, in 2008 he correctly predicted the presidential results in 49 out of 50 states. And he's on pace to be 50 out of 50 if Florida closes the way most experts think it will.

You can also imagine how many heavily partisan "pundits" attacked him in the lead-up to Election Day. If Silver's numbers didn't favor their party’s candidate (and even in some cases where they did), those paid to opine on political topics discounted his mathematical approach and derided him personally. It actually created quite a “moment of truth” on election eve – were the experts right, or would math win the day? We know now that math won.

So what does this mean to us as digital marketers? This again reinforces the power of a data-driven mindset. In the face of massive amounts of skewed and contradictory data, Silver did what great marketers do – develop a model. We do it every single day, except we call it segmentation or A/B testing. It’s the same exercise, and we’re all at different points in the progression of becoming a data-driven marketer.

Science + Art = Greatness
Before I begin to sound like data is the single Holy Grail of marketing, let’s take a minute to review a few best practices. I have always contended – and most people agree – that great marketing is equal parts science and art. The science is best applied to uncover unseen insights in your customers’ demographic information and response data. We can tie together actions to create hypotheses about who will respond, what they’ll spend, or how they’ll act.

But the numbers are just that – numbers. The art of marketing comes in the offer development and creative execution. You might spend 25 hours isolating the most pristine list of engaged customers, but if you send them a weak offer or an unclear call to action then you’ve utterly wasted all that time and effort.

The highest performing marketing groups I work with have equal parts science and art. I’m seeing more Web developers integrated into marketing teams so that landing pages are optimized for SEO and lead tracking from the start. I also see dotted-line reporting of database analysts to marketing managers in order to drive pre-campaign audience sizing and “what-if” scenario building. And just recently I’ve seen a traditional publisher with more than 50 titles building a propensity model to understand its subscribers’ potential affinity for other related titles. The possibilities become almost endless once you embrace the science + art mindset.

Not Really “Big Data”
As a side note, almost everyone is talking about how Silver is using “Big Data,” which in this case isn't 100 percent accurate. True big data combines massive amounts of raw data to allow unstructured queries that produce unique insights. In this case, Silver is rolling up already-massaged data and creating a super-set analysis designed to alleviate bias. And I know more than a few people who think any calculation that can be run on less than 10 parallel machines isn’t really big data anyway. But we’ll save that argument for another post.

For more on big data and how it will impact marketing in 2013, download Silverpop’s new white paper, “6 Key Marketing Trends for 2013 – and Tips for Succeeding in the Year of the Customer.”

Related Resources:
1) Infographic: “Comparing Barack Obama and Mitt Romney’s Emails, Post-Convention
2) Blog: “Why Gary Vaynerchuk Should Be Your Marketing Superhero
3) Blog: “Using Real-Time Data and Relational Tables to Drive Customer Emails



Sign up Now!

Subscribe to IBM Marketing Cloud's Digital Marketer Newsletter!

Popular Categories

Top 5 Posts


To give you the best experience, this website uses cookies.

Continuing to use this website means that you consent to our using cookies. You can change your cookie settings in your browser at any time.
Find out more here or by clicking the Cookie Policy link at the bottom of this page.