Moneyball and Programmatic Ad Buying Algorithms

Programmatic Ad Buying, buying algorithms

The story of Moneyball is familiar to many. Advance Scout Billy Beane used a system of data on baseball players to pull together an inexpensive but quality team roster for the Oakland Athletics. This system was first developed by Bill James in 1977, who had created an 80-page book about hits, runs, pitching and more. James and others like Beane saw the value in this data and began digging into it to find important insights.

The goal of this type of data mining was to find the “diamond in the rough;” players that could offer great value to teams, but who were also underrated (and therefore less expensive to sign). The larger goals of this type of search are the same goals of programmatic audience buying.

Moneyball and Programmatic Ad Buying Algorithms

In the Moneyball algorithm, the goal was to use data analysis to find valuable players. The objective, however, was not to only focus on batting averages and similar statistics. Instead, it was to try to predict how potentially beneficial a player could be based on more in-depth, lesser-known statistics. These stats could help teams make better decisions and create better outcomes.

Programmatic buying relies on data to make similar decisions for campaigns. It looks at granular data to determine target audiences and what campaign elements are actually being effective. Similar to only looking at batting averages, many marketers only look at clicks to determine campaign success.

The problem with this, however, is that many systems – especially legacy ad tracking platforms – only focus on last-click conversion. This does not give credit to other campaign tactics that might have contributed to the conversions, i.e. brand awareness (display or social media) or email campaigns.

By looking into the data, marketers can see how these “lesser-valued” campaigns and data points may actually be contributing to positive ROI. Basically, each “player” within the campaign can be given credit throughout the conversion funnel. In the Moneyball algorithm analogy, value is given to every player in the funnel as opposed to the major hitter/last click converter.

Takeaways

Digging into all of the campaign data helps marketers create a comprehensive strategy to develop winning campaigns. Lesser-valued marketing players can actually be important and contributing to the overall goal of conversions.
Before eliminating a tactic, review the data and the insights that they provide. Oftentimes, tactics are working in tandem to create a comprehensive whole – like the Oakland A’s team in Moneyball. A lesser-valued, inexpensive player could actually be the key to a successful franchise.