Turning prediction into an analytics game

Predictive analytics is about obtaining a deeper understanding of the customer and their environment, as well as delivering future-focused insights.

Johannesburg, 30 Jul 2019
Read time 4min 10sec
Dario Debarbieri, CEO for Asia Pacific, Enterprise Outsourcing.
Dario Debarbieri, CEO for Asia Pacific, Enterprise Outsourcing.

We are witnessing a faster evolution of technology today than at any other time in history. This has led to our ability to generate content and data at high speed. This data explosion and the concomitant digitisation of information has left organisations, for all intents and purposes, swimming in an ocean of data.

According to Dario Debarbieri, Enterprise Outsourcing’s CEO for Asia Pacific, most businesses are aware of the value of information and make it a point to obtain customer data. However, such data is generally restricted to the organisation’s interaction with the individual client.

“Naturally, if an enterprise wants to obtain a holistic view of the customer, they also need to consider information from elsewhere, such as that found on social media. Remember that although a customer may tell your business they are happy with your service, they may say the opposite on social media. So, using analytics across these various data streams to obtain a more complete view of the customer is crucial,” he says.

“The next step is to begin utilising predictive analytics, which is about obtaining an even deeper understanding of the customer and their environment, as well as delivering future-focused insights. Whereas standard analytics deal with historical data, predictive analytics offers algorithms to use historical and current data to inform your views of the future as well.”

As an example, he points to a scenario where a building manager has to take care of a number of lifts. Being machines, their parts will eventually begin to degrade. But predictive analytics will enable the manager to know when these parts are likely to break, so preventative maintenance can be scheduled. The advantages are numerous: not only will this reduce maintenance times, which can also be staggered, but it also improves user safety, increases users’ satisfaction and saves the company money.

“Predictive analytics can also help organisations identify inefficiencies in their systems and thus take steps to eliminate these. If a restaurant, for example, is able to predict when it will run out of a particular ingredient, more can be sourced prior to this, thereby ensuring business continuity,” says Debarbieri.

“Other advantages offered by predictive analytics include reducing costs, enabling businesses to anticipate events, reductions in downtime and, of course, increased revenue. Moreover, depending on the type of business, it may also improve things like asset management and, ultimately, enable the company to make more intelligent decisions.”

Debarbieri adds that a wonderful real-world example of predictive analytics in action are the modern map applications: they are able to predict the best route for a driver and constantly update in real-time. These updates, however, will only be helpful if the application has access to as much data about the area being travelled as possible; then they will be able to save the user fuel and time, and ultimately, boost their efficiency.

“A major impediment to successfully implementing predictive analytics,” he continues, “is simply gaining access to enough relevant data. Obviously, the more data you have that is accurate, the better your prediction model will be. And accuracy is critical because if the data is inaccurate, the predictions will inevitably be wrong, since if you put garbage in, you will get garbage out.

“It must be noted, however, that while there is a lot of useful data in the public domain, when it comes to using customer information in this way, enterprises will simply have to have a much higher degree of integrity around the protection of this data. I would suggest that all customer information be treated as if it was the company’s own intellectual property.  In fact, the way customer information is treated should be the same as if you were dealing with your client face to face.”

Debarbieri says businesses must also remember that they need to continually invest in the relationship. This means using some of the money saved through the benefits of adopting such technology to invest in more effectively securing the client data and making sure it is as accurate as possible.

“In the modern world, it is obvious that data is king, and it is aptly described as the ‘new oil’. This is, undoubtedly, because of the effect it can have on both businesses and their customers. Moreover, while data is the foundation on which predictive analytics is built, such analytics are, in turn, the foundation on which true artificial intelligence (AI) will be born,” he concludes.