Don't just skate to the puck

As I walked past Madison Square Garden on my way to another meeting here in the Big Apple, I started to think about the start of the NHL hockey season and the timeless metaphor that businesses all need to act like Wayne Gretzky and skate to where the puck is going not where it currently is on the ice. 

When I was an executive at Cognizant, this was one of CEO Frank D'Souza's favorite analogies (even when he was addressing employees in India where most had never experienced ice hockey). Frank used it to highlight the need for the right business strategy, but today you must think of it in the context of data and analytics, because data and analytics is your business. Here are the four steps we must take so you not only skate to the puck on a single play, but prescribe a formula to win the Stanley Cup:

1. First you must prepare. In hockey (like most sports) this means practice, practice, practice. By running every play over and over you learn what is possible. In data and analytics you need the right strategy. What data and business requirement is right for a Hadoop cluster and what business data can remain in legacy data warehouses? What data has competitive advantage and what data does not? What data sources are reliable and what data sources need continuous cleansing and munging? Spend the time on the strategy, it is the business imperative.
2. Second you must understand and be able to describe what happened in the last game and how you react and adjust the way you play in tonight's game. If your hockey opponent scored last night using a certain formation in front of the crease (goal), you can bet tonight's opponent will run the same play if you do not adjust. In data and analytics this is descriptive analytics. And while everyone talks about the future of self drive everything, it is remarkable how many companies screw up and over complicate the reporting of what happened yesterday. Proper data cleansing and munging is imperative otherwise it is garbage in which means garbage out. No rocket science needed here and the tools can be as simple as an Excel spreadsheet, coupled with a clean, precise formula and process exercised flawlessly is the key.
3. Third comes the Wayne Gretzky analogy of skating to where the puck is going. You've studied the game film and anticipated how your opponent will adjust their play when they watch the film of your game last night. You see Messier at the point and before he raises his stick you are heading to the slot so you can redirect the puck and score. In data and analytics you have analyzed the last five trade promotions and modeled them against public sources like weather, social sentiment and other factors and you run the promotion that predicts the consumer's behavior. Your sales increase and your profits grow as you optimize inventory and distribution for the outcome you predicted.
4. Fourth is how you become a Stanley Cup Champion! You create a plan that continually adjusts to not simply predict a single play in one game, but a self prescribing flow of plays and adjustments that help you outperform your opponent for the 82 game regular season and most importantly the 16 playoff wins required to hoist the Cup. In the business of data and analytics you combine business rules, algorithms, machine learning (the marketers call it Artificial Intelligence or AI) and computational modeling procedures to model not each individual scenario, but how they change based on dynamic scenarios and sequencing. Not every market scenario can be predicted, but you can help to prescribe the market scenario so the outcome to your business is favorable. This is where you do need the rocket scientists.

To summarize... In hockey you practice your play, learn from last night's game so you can predict how your opponent will adjust tonight and then prescribe the right sequence of future plays to win continuously over a long season. In the business of analytics, start with the right data and analytics strategy, learn to describe what happened, learn to predict the next single event so you can ultimately have the ability to prescribe the future outcome across a broad sequence of events.

Okay on to my next meeting...