Predictive Analytics

Artificial intelligence (AI) and predictive analytics are becoming more important for more businesses today, but if you are not familiar with the mathematical models how do you understand the opportunities? What we are really exploring is the power to predict what your customers will do in the future, based on their previous behavior and other variables.

That sounds simple, but what are some real-world examples of how this can work?

Think about a bank with millions of credit card customers. The collections function at the bank is responsible for chasing customers who don’t pay their bills. There will be a significant investment in this process, in addition to the problem of not actually receiving the cash on time. What if you could use predictive analytics to look at customer behavior and predict which customers are going to have a problem paying on time this month?

If you can step in and offer help to the customer before they default then it’s likely that you can not only ensure that the payment will happen, but you have also dramatically improved the customer experience – that customer will feel far more loyal to a bank that helps them out.

There is a similar effect on subscription businesses too. This is when the customer is paying a fee every month to access a service. Maybe it’s a magazine or Netflix, or their mobile phone contract. There will often be certain behaviors displayed by customers that are about to leave their subscription. Perhaps their movie watching dramatically declines as they start using a rival service or their phone use ceases when they were previously a regular user.

In this case, there is the opportunity for a company to step in and offer a deal or package that will convince the customer to stay, even though the predictive analytics has identified that this customer is highly likely to stop using the service. If Netflix calls and says ‘Would you commit to another year with us if we give you a 20% discount?’ many customers are likely to respond to the offer.

We are seeing AI deployed in many other areas, such as intelligent automation and chatbots, and many of these solutions do work well, but there have been some disastrous implementations that make many executives wary.

I would step back and separate the intelligence needed for predictive analytics from the more general AI used in direct customer interactions. You can use analytics to develop and train your internal team by using the system to identify weaknesses and areas that need improvement. Not so you can exert more control, but so you can support the team and help them to improve faster.

Every company would love to predict what their customers will do next. Predictive analytics can give you the partial ability to do this. Customers will sometimes take random actions, but in most cases, you can apply their previous behavior to an algorithm that also factors in environmental variables – the weather today, the traffic, special events taking place.

Retailers can form a much more supportive relationship with customers by knowing what they need and when. It can even create proactive opportunities where brands can suggest ideas to their customers. How many more cold beers could a retailer sell if they reach out to fans of beer automatically based on their previous spending patterns and the knowledge that a major football match will take place in the next few days? Predicting demand can create opportunities.

I believe that all companies across all industries will be using these tools to get closer to their customers. Brands that are not exploring the data they have will look extremely old-fashioned and will eventually fail. Who could now imagine what is special about a  discount if it applies to every customer?

Using special tools and systems, one can optimize business processes and predict customers’ behavior. Read more about the intersection between technology and business solutions, and discover how predictive analytics can help your business.

.