No doubt you have at one time or another called a customer service line and were met with a voicemail tree assigning options to various numbers. Often, and amazingly, the reason you’re calling is not listed among the options. That company’s predictive analytics came up short.
After keeping track of and studying why most people call, the brand whose help you need determined the top options their customers need on the service line. You just fell through the cracks. You have to select “Other.” A likely but even more disturbing possibility is that the brand didn’t study their data at all and simply guessed amongst themselves what options customers would need.
Misfires like that are more impactful now than they used to be thanks to the empowered consumer. They willingly give organizations an enormous amount of personal and consumer data, from a variety of sources and in a variety of ways. In return, they fully expect brands to use that data intelligently to know what customers want, and to provide seamless customer sales and service experiences no matter what the touch point.
Customers really notice when you don’t know them and don’t appear to care about them. They don’t like it, and it’s a short hop to them not liking you.
Predictive analytics not only shows that you know them, it shows you care enough about them to know in advance what they will like or need. It’s not terribly unlike someone who can predict with great accuracy what their significant other will like for a birthday present. Get it wrong and you’re sleeping on the couch.
Predictive Analytics World runs down the definition as “combing through past info to derive models and analyses that help project future outcomes.” It can be used to learn what customers want and learn ways to optimize operations. That means efficiencies, cost savings and happy consumers. What makes that possible? Something we talk about a lot these days, big data.
We’ve also talked a lot about the merging roles of the CMO and CIO. Traditionally, data has been the domain of IT. But increasingly, marketing is being held accountable for the measurable, effective use of that data, which includes predictive analytics. The volume of data and the speed at which it comes in is now an enterprise reality that’s forcing historic structural and operational changes internally. Predictive analytics can intelligently inform those changes.
Managed properly, real-time actions, reactions, and changes in the marketplace can now be added to historical data in models. Managed poorly, or if predictive analytics models are requested but then not deployed successfully in the business, potential benefits are lost. Best Buy learned under 10% of customers constituted almost 45% of its sales and redesigned stores accordingly. Predictive analytics are used to forecast how patients will feel about drugs and treatments. And you’ve probably seen news reports about how city police departments are using predictive analytics to prevent crime before it happens (cue comparisons to “Minority Report.”)
If you determine that customer acquisition, customer sales, customer service, customer retention, customer reactivation, and operational efficiencies are relevant to your organization, it’s time to make sure your social engagement and monitoring tool is pulling in the social data that can be integrated with enterprise data so that quality predictive analytics models can be run.
Photo: James Barker, freedigitalphotos.net