Automation and algorithms are increasingly dominating financial processes. One important technological development is predictive analytics. What is it, what makes it successful, and what can – or should – the CFO do with it?
The next step
Predictive analytics is not new. Large companies have been using it for many years for various issues with regard to risk reduction. Consider the banking sector or the insurance industry. They use this technology eg. to decide how high premiums should be, or whether or not to issue credit. Predictive analytics offers more potential now than ever due to the fact that we have more internal and external data at our disposal. It enables us to delve into these enormous databases and to investigate patterns and structures. These insights allow us to predict future results.
As a stage in the evolution of analytics, predictive analytics follows descriptive and diagnostic analytics, according to research agency Gartner. With descriptive and diagnosticanalysis , primarily historical data was examined to explain why certain things happened as they did. Predictive analysis goes further than that, using data as a crystal ball for predicting the future. As CFO, this enables you to base your decision-making on hard figures instead of your intuition. What’s more, Gartner’s research shows that only 13% of companies use predictive analysis for their current strategic decision-making.
Predictive analytics technology is now seeping into other industries, because companies have collected large amounts of suitable data in so many different areas. Additionally, an increasing number of tools are becoming available which facilitates analysis based on various data sources.
Good practice: Netflix
Netflix is a good example of a company that successfully uses predictive analysis. They collect data on their customers’ viewing behaviour and combine that with data from external databases. On the basis of this combination they can successfully determine how to retain their customers and where new growth markets are emerging. Consider data from the rating system, search queries, stop/pause data, and data from (mobile) devices that people use to watch films and series.
And let’s not forget House of Cards. Every decision which helped make this a hit series was based on hard facts. No guesswork, no intuition. From cast members to the use of colour, every decision was made on the basis of predictive analytics. And what about its use of ten different trailers, allowing viewers to be shown the version which suits their specific viewing habits? This is predictive analysis on an entirely different level.
The next stage in the development of analytics is prescriptive analytics. With this new form, we see a number of interesting additions: neurological structures, machine learning, graphics analysis, simulations and the processing of complex events. The system provides predictions on the basis of data from these different sources and makes really interesting suggestions. Given the rapid developments in these areas, it won’t take long for that phase to arrive.
What can – or should - CFOs do with predictive analytics?
Until now, we’ve been limited to the descriptive and diagnostic analysis phase. And predictive analysis seems like a Fata Morgana. We are willing to embrace it, but we don’t know what to do with it yet. This is something many companies are grappling with, which is a shame, because every company still struggles to stay within budget. With predictive analyses, you can determine budgets more effectively. The more diverse your data sources, the better your forecasts will be.
This obviously requires a new setup of the finance function in general and of the CFO role in particular. The CFO has to become a digital leader and say goodbye to the focus on control. The CFO will need to focus on the future of predictive analyses on the basis of accurate and effective reports. Only then can you become a real strategic partner for your company. When you are able to predict customer behaviour, you’ll be able to advise your company effectively, leading to better results, innovations, constant growth and a grip on risks.
Establish the right technology
How can predictive analysis be well implemented? And how can data be used effectively? There are three issues commonly confronting the CFO in this task. The first is that he or she often has no access to a database which combines internal and external sources. Secondly, good algorithms are often lacking. Finally, the results of the analysis are not presented clearly enough. This means that the CFO still doesn’t have the insight necessary for good strategic decision-making. Which sounds strange, because there are more business data tools around now than ever. But that’s a matter of quantity versus quality: you might have numerous tools at your disposal, but it’s crucial to deploy the correct one. Moreover, CFOs must have a clear vision regarding the use of data analysis and on the interpretation of that data.
Ask the right questions and don’t wait too long
Everything starts with the combination and connection of various relevant data sources. This immediately raises questions. How can this be set up and integrated correctly? What data is relevant and what isn’t? And particularly of concern for CFOs: Do you hire specialists for the data analysis and the setup of such a data warehouse? Are these people part of your team or do they fit better elsewhere in the organisation?
These are all good questions. But don’t wait too long, because things are developing at a rapid pace. Ultimately, the goal is to prevent your competitors from overtaking you, and from having to to admit: ‘I didn’t see that coming…’. As CFO, you want to fulfil the role of strategic partner and be proactive. That’s why it’s essential to embrace and deploy predictive analytics. That’s how you gain crucial insight into the future, based on facts.