The Benefits of Predictive Customer Analytics
Predictive analytics can provide highly accurate predictions that allow for more effective business decisions and product investments.
By Scott Sanders
A company’s ability to understand its customers’ behavior and the nature and speed of how those behaviors change can make it or break it. For a Chief Marketing Officer (CMO), nothing is more imperative to smart decision making and strategic planning.
To date, the most common kind of customer analytics that the CMO receives is a result of measuring and analyzing Website traffic, click-throughs, disparate CRM information, customer satisfaction and attrition. However, these descriptive analytics, which describe what has happened in the past, are nearly obsolete by the time the CMO gets them and, therefore, only provide limited value. You cannot retain and upsell customers who have already left your company. The higher value from IT is to enable the CMO to leverage predictive customer analytics.
Predictive analytics do not tell us precisely what will happen in the future. Instead, statistical modeling and machine learning help identify patterns and trends that can be used to calculate the probability of future outcomes. For example, predictive analytics will not tell us in advance which single NASDAQ stock will be the biggest winner in 2015 or which NFL team will win the Super Bowl in 2016. Both of these scenarios have too many unknown variables at play.
Furthermore, the future outcome probabilities defined by predictive analytics are by no means fool-proof. Remember the models that were developed for investors to predict risk profiles of subprime mortgages not too many years ago? Likewise, one cannot expect that predictive analytics will tell retail store marketers exactly which promotional idea will drive the most footfalls per dollar spent.
However, when modeled correctly, predictive analytics can provide highly accurate predictions that allow for more effective business decisions and product investments. By identifying patterns in a customer’s buying habits, for example, Target can predict which of its female customers are most likely to become pregnant and their approximate delivery date. With this information, the company can send targeted coupons to both her and the immediate members of her family for baby-related products, increasing the likelihood of a purchase.
Most CMOs, by nature, are cerebral “creatives” with entrepreneurial business acumen and intuition. Experienced CMOs make it a priority to understand their customers’ buying habits, the competition and the ever-changing demographics that affect the market. After all, developing campaigns and products to help drive sales growth and brand recognition is at the heart of their job.
Adding predictive customer analytics capability to the CMO’s toolbox can put his or her business intuition on steroids. Imagine the CMO of a consumer products company who is able to predict the behavior of individual customers or which products that customer is most likely to buy next. There is no question that a hotel chain with statistical analysis showing how many room-nights a customer will buy over his lifetime or a bank that can predict commercial accounts attrition in time to take preventative action will outperform the competition by leaps and bounds.
They understand that consumer buying decisions are full of knowns and unknowns. CMOs simply want data that is easier to access, more timely and more telling. IT should make it a priority to enable these predictive customer analytics, helping the CMO implement standardized processes for mining customer data and leverage that information through the use of statistical modeling and machine learning. If the required skillset does not exist internally, a new breed of data science consulting firms are available to help design and build systems that capture and report historical and real-time customer data. A handful of third-party tools and services are also on the market to keep the process running.
Enterprises that disregard this trend will do so at their own peril. If IT departments haven’t already heard the clamoring from the CMO for more predictive data, they soon will. The chain reaction happens this way: sales will complain that it can’t do its job effectively because marketing is ineffective, and marketing will say it could perform better if only it could get some data out of IT that isn’t a year old or that isn’t scattered about in multiple places indiscernible to the naked eye. Add to this the growing trend of Fortune 1000 companies promoting CMOs to oversee both marketing and sales staffs, and IT will start feeling pressure from the CMO all the more.
Scott Sanders is the director of Information Services Group (ISG).