With predictive analytics, you can take a look into your company’s future. So for instance you can make predictions on your most lucrative prospects, market evolutions, bad debts and fraud. Chief data scientist Joris Peeters explains all about the latest developments and insights, and peers into the crystal ball of predictive analytics.
Nowadays cars drive themselves, you can print 3D objects at home and we can operate devices with our thoughts. So it’s not so strange that we’re able to predict what the future holds, is it?
The magic formula, according to Altares/Dun & Bradstreet’s chief data scientist Joris Peeters, is: an ever-increasing mountain of data + data mining + machine-learning technology + statistical algorithms and models.
Peeters says that this formula can produce state-of-the-art predictions. “For example, our models can predict three-quarters of the bankruptcies that will occur in the Benelux in the next 12 months.”
Infographic: prosper from predictive analytics
The potential applications of predictive analytics for businesses are countless, says Peeters. “Prospecting, compliance, credit scoring, process automation, credit management, logistics and customer analytics: you can actually use predictive modeling data and technology at every stage of the business cycle.”
Let’s briefly zoom in on prospecting. Peeters says: By giving a simple command to a database, you can separate quite a lot of the wheat from the chaff, for instance by filtering out the smallest companies. But the biggest benefits come when you apply predictive modeling. This will let you know which prospects have the highest potential and should be contacted first. Imagine that 2 percent of the companies that you cold call usually end up buying from you. With predictive modeling you can raise that percentage substantially, perhaps to as much as 15 or 20 percent.”
There are countless ways in which predictive analytics can help companies to flourish: by identifying which customers might leave them, predicting which suppliers are going to get into difficulty, discovering in good time which transactions pose a heightened risk of fraud or corruption, to name but a few examples. Would you like to find out more? In that case, read our white paper: “Turning data into insights: predictive analytics”.
Analytics: focus areas for companies
Joris Peeters says that companies in the Benelux are rapidly realizing that they need to get serious about business analytics and predictive analytics. However, to turn data into insights, you need a lot of knowledge in your organization. What does Peeters think that companies should focus on most?
1. Understand the potential of data
Companies don’t always understand what data can do for them. Peeters says: “If you don’t constantly investigate what’s possible, you’ll fall further and further behind. Those companies whose managers understand the knowledge that can be derived from data will be the big winners in the future. We’re already seeing a growing divide in the market.”
2. Invest in knowledge and in the right people
Companies require a lot of knowledge on data and analytics. But who should you hire? “I’ve seen companies with data science teams comprised only of math graduates. But their return on investment was pitiful. Those people may know all there is to know about statistics, but they don’t have much business sense,” says Peeters. His advice is: the people at C-level need to have sufficient knowledge of what data can do, and ideas about how to use it. Lower in the organization, there needs to be a strong link between data science and the commercial side of operations.
3. Don’t invest blindly in hardware and software
Technology should be an enabler, not a goal in itself, according to Peeters. Too many companies become obsessed with particular hardware or software, without asking themselves what they want to do with the data. “Start by establishing what the purpose of the data is; only then should you ask yourself which technology you need,” the data scientist says. Ultimately it’s all about three things working in harmony: appropriate technology, good data and competent people.
4. Contributors to evolution
As a trailblazer in predictive modeling for businesses, Altares/Dun & Bradstreet is set to make a considerable contribution to the evolution of predictive analytics in the coming decade. Peeters says that he and his colleagues will have to deal with three major developments:
How to link the ever-increasing mountain of unstructured data, for instance from social media, back to the companies that it concerns? Artificial intelligence will have to play a substantial role in these efforts. And that’s work in progress.
Companies quite rightly expect the models and platforms to keep improving. They expect the predictions to become more accurate and that they’ll be able to seamlessly integrate our off-the-shelf predictive analytics, for instance from D&B Credit, D&B Onboard and Market Insight into all their workflows.”
And there’s going to be more and more focus on mobile devices. Customers no longer want to have to read 15-page reports. Simplicity, speed, convenience and efficiency are key. And that means moving towards mobile.
5. Predictions about predictive analytics
Where does Peeters believe that we’ll be with predictive analytics in 10 years’ time? Peeters says: “We’re going to see spectacular developments at every stage of the business cycle. For instance, we’re now also collecting data on transport. In the future, haulage companies who’ve got some space in their trucks will be able to check an app to find out whether they can pick up extra shipments from other companies.”
Peering into his crystal ball, Peeters also says that working days are going to be completely transformed. “Imagine, you’re working as a sales rep. An intelligent system will draw up a list of prospects for you on your phone. You’ll order a self-driving car, plug in your phone and the car will automatically take the smartest route to all of those prospects. When you visit a prospect, you’ll be able to carry out all kinds of checks on your phone right away, for instance on credit scores and compliance. On your way home you might send a few fully-automated invoices.”
It all sounds rather futuristic, doesn’t it? Peeters says: “Not as much as you’d think. We’ve already discussed the sales scenario I described with customers. Believe me, it’s closer than you might realize.”
Want to find out more about the applications of predictive analytics, focus areas for companies, anticipatory analytics, business analytics trends and the predictive models offered by Altares/Dun & Bradstreet? Then download our white paper:“Turning data into insights: predictive analytics”.