By Dheeraj Pandey, IQR Consulting
While shopping on a major retailer’s web site recently, I abandoned the shopping cart with one item remaining in the cart. Approximately two hours later I received an email with a coupon worth 20% off any item. Coincidence? Not likely.
Many retailers, partly driven by the success of Amazon.com, are tapping into the data supplied through their web sites to create customized advertising and promotional offers. And with the proliferation of smartphones and GPS, the use of that data is beginning to translate to the brick-and-mortar retailers.
How powerful would it be to your bottom line to be able to predict a specific customer’s future purchases then push out an offer or incentive via text message based on that information the moment that customer walks through the door of your business? Believe it or not, technology has already produced the capability to do that and it’s all because of something called predictive analysis.
If you’ve spent any time in a sales or marketing meeting, you know that business has traditionally analyzed data based on what has happened in the past, used the data to report results, and then set strategy based on those results. Predictive analysis takes that process one step further by using those same known outcomes based on an analysis of historical data, applying certain variables to those outcomes, and then predicting the likelihood of a similar scenario occurring again. Through the use of various techniques including statistics (yes, there are people who thrive in those dreaded college statistics classes), modeling and data mining, analysts can evaluate data and find patterns within the data that indicate consumer trends.
Predictive analytics is not new and has been used in the insurance industry for years as a way to manage risk. Insurance carriers have established policy premiums by using actuarial tables to evaluate the likelihood of mortality. Mortality data can be based on age, gender, health and other variables that are then used to establish a set of risk classes based on the likelihood of the carrier having to pay a death benefit. These risk classes are then used to determine an individual’s premium.
The success of predictive analysis still depends on the assumptions that were drawn from the data, but the future looks promising for retailers. Whether it’s to better understand and act upon customer behavior, to evaluate the probability of fraud or predict the success rate of new loyalty programs, predictive analytics can determine the “So what?” of data by finding the patterns that drive informed results so that retailers can develop more responsive and personalized marketing strategies. And the capabilities will only grow from here.
As technology continues to improve at a dizzying pace, data analytics is sure to improve with it, as will the results that are derived from consistent and thoughtful analysis. The companies that embrace predictive analytics will have more precise customer intelligence leading to targeted consumer offers and, ultimately, happier and more loyal customers.
Dheeraj Pandey is a Senior Consultant at IQR Consulting, a provider of data analytics solutions to retail and financial services firms. He provides analytics, business intelligence and quantitative research for the retail and financial industry, and his research on consumer behavior has been selected for presentation at several international conferences, including INFORMS (The Institute for Operations Research and the Management Sciences). Pandey’s main areas of interest include behavioral sciences, data modeling and social media marketing.