NCAA Tournament Mirrors DCO Strategy

My NCAA bracket is hanging on by a thread and all my hopes and dreams (and $20) are pinned on Kentucky!  I don’t know too much about college basketball because I went to a football school (Roll Tide!), so I made my picks based on teams I like (or for that matter, don’t like), teams my friends like and teams I know only come out of the woodwork in March like Gonzaga, Marquette and Xavier.  Needless to say, my bracket is based more on emotion and limited knowledge, than research and statistics.  But all of this Madness got me thinking about how basketball is very similar to Dynamic Content Optimization (DCO).  I know, I know, it seems like a stretch, but it’s not – hear me out…

Before a game, the coach designs a game plan and forms his (or her) lineup. There are some givens a coach knows going into a game – the different positions and players on the team, specific players’ strengths and weaknesses and how the players have been performing during practice.  A coach also studies their opponents, analyzing their strengths, weaknesses and strategy.  This is all very similar to the pre-optimization or rules-based learnings of a DCO campaign – taking what you know about your business, your advertising strategy, your audience and your specific campaigns and mapping out specific rules for your dynamic campaign.  You might know that you want to serve a specific product, message or offer to one audience, whereas another audience should receive a different product, message or offer.  There are a number of variables or inputs that you can use in an ad – backgrounds, images, products, offers, messaging, calls to action, videos, and much, much more.  You can utilize pre-optimization, given what you already know about your campaign to make sure that each of the inputs are served to the right person at the right time,  thus creating multiple creative versions in a snap!  No need for creating hundreds of creative units – we start with a shell and dynamically layer in each input to create the right output or creative unit, all in real-time!

So you do your best to prepare and plan your strategy before the game, but after tip-off, everything might change.  You realize that some tweaks need to be made to your line up and your strategy.  Maybe your point guard isn’t really on point or your three point ace hasn’t sunk one yet, and while you thought your opponent was going to play man to man, they show up playing zone.  All of the sudden you realize that you have to optimize your game, your players and your strategy.  You make some quick, game-time decisions to put in different players, run alternate plays and change up your game plan.  This is similar to machine-based learning and optimization in DCO.  Based on performance against specific metrics and goals, the auto-optimization engine combined with creative versioning allows the server to serve up different elements or inputs to output different creative version to yield maximum results for your campaign.  This is all done on the fly and in real-time.

As you might imagine, one type of optimization is no good without the other.  If you didn’t do your homework or study your team and relied solely on game-time decisions, you likely wouldn’t even make it to the Big Dance.  On the other hand, if you did your due diligence, spent countless hours forming the best line-up and watched obscene hours of your opponents’ game films, but left your whiteboard at home during the game and sat quietly on the bench and changed nothing as you team and game fell apart,  even the NIT would be a pipe dream.  It’s important to understand your team (your brand), know and study your opponents (your competition) and then be agile and responsive during the game (your campaign) and dynamically optimize for a big win (even if it is a buzzer beater)!