Corporate Performance Management Needs Data Science

These days, corporate Performance Management (CPM) is not just about KPIs, and reporting on what has already happened. CEOs and Shareholders are much more interested in what is ahead and how their business is going to look in the next Quarters, Years and Decades.

If anyone has attempted to forecast their business beyond the end of the financial year, then you will know that a simple extrapolation of the past just doesn't work. External influences within your industry, your country and the world do not affect your business much in the near term, but beyond 1 year it is crucial to understand the impact of the world around you on your bottom line.

For this, you need Data Science. Let's use an example. Say you are a retail business and you need to forecast your profits for the next 5 years. In this environment, profits are driven by the behaviour of your millions of individual customers. In order to predict how these customers will behave in the future, you will look at how they have behaved in the past. Before you can do this however, you have to group them together. Now traditionally, retail businesses tend to group their customers by income, because the assumption is that if you are in a certain salary bracket, then you can afford certain products. This is a good starting point, as its probably reasonable to assume that people of similar income behave in a similar way, but where do you draw the line between one income bucket and the next, and how do you know this is the best place to draw the line? Surely this will change over time with inflation, exchange rates, etc?

Why not let the computer decide how to group customers? Give it all the data you have, including the fact that customers of various income groups are spending a lot of money on nappies, and others on overseas trips. The computer can find commonalities across customers with combinations of behaviours that a human will never be able to detect. This is called machine learning, and it's a mission critical tool for some of the fastest growing companies on the planet today (think Google, Apple, Uber, AirBnB).

Once you can group your customers together in a way that allows you to predict their behaviour as a group more accurately and further into the future, then you can start playing around with what drives these behaviours (internal factors such as pricing, and external factors such as the economy) and get a much better understanding of how to get more of the great customers and less of the bad ones. Do you think this will have an impact on your bottom line? Most definitely!

The field of Data Science combines Mathematics and Statistics, Computer Science and Domain Knowledge to provide customer insights that just weren't possible a few years ago due to the power of computer hardware. There are a range of new tools out there that take advantage of new hardware to allow the average company to use Data Science and Data Scientists to become the Exceptional Company.