Programmatic Moneyball

Posted by Ted Yang on June 11th, 2014 at 9:07 am

In the early days of online advertising, if advertisers wanted to make a splash with their brand, they would buy a full-page digital takeover on a big name publisher’s website. Advertisers would turn to the biggest possible publisher hoping that their impact on a well-known site would drive results.

With notable exceptions, programmatic has changed this game. Advertisers still use big-name publishers, but now they can make decisions based on the numbers that actually matter.

It’s similar to what happened in Michael Lewis’s book, Moneyball. In the book, Lewis recounts the tale of how Oakland A’s manager Billy Beane took an analytical approach to building a successful team that didn’t focus on big names or typical home run production. He built a winning team in a much more cost effective way by focusing on what was analytically proven to be effective, like on base percentage – not just what people thought was effective like home runs.

Beane’s approach was to hire players that were statistically more likely to contribute to the kind of success that actually wins games. Rather than look at the size and shape of an athlete, Beane looked at how well players played on the team in coordination with other players. Rather than picking the safe names that baseball wisdom had traditionally valued, he focused on rigorous statistical analysis and realized that time spent on base and slugging averages were better success indicators than RBIs and batting averages. The team used its limited budget to hire players that statistically played well together, rather than focusing on big name reputation. Overall, the team spent less money, but still managed to win a lot of games.

If you have the payroll of the Yankees, you can afford to go out and buy the best players without looking too closely at the numbers. But most marketers in this day and age, even the largest companies, can’t afford to play like that. So instead, they take advantage of programmatic digital media to figure out the data points that actually matter when deciding where to place their ads, how often and to whom.

In the programmatic world, it is all about gathering the right information, analyzing that information and then creating a feedback loop for immediate optimization. Rather than selecting a publisher because of reputation or a gut feeling, advertisers can validate their choices by looking at the actual versus forecast results, then adapt...quickly. Advertisers shouldn’t rule out the big names; they should put big names in contention and measure their results just like everyone else. The Moneyball approach means not relying on expensive big name players who everyone else thinks are good if they are not the right ones for your team. Beane’s whole point was that he wanted teams that were going to win games regardless of reputation. Get three players who work well together for less money than hiring one all-star.

It’s not that baseball didn’t have stats before the Oakland A’s took this approach. To the contrary, die-hard baseball fans can rattle off more stats than most quants! But what Billy Beane did was figure out how to use those stats to the team’s advantage in a new way.

The same can be said for digital media. It’s not that digital didn’t have stats before programmatic, it’s just that before programmatic, there was no actionable way to use those stats.

Advertisers who are looking to make a big splash should consider their options and not assume that a big name will win the game. Nowadays, it’s a numbers game, a stats game. The brand with the best approach to the numbers will win games today and in the future at less risk.

Leave a comment