Analytics means different things todifferent people. Some think it is a department that helps inform marketingwith A/B testing (test and control set up and evaluation), targeting, andmeasurement. Others see it as a practice that falls within IT to producereports and dashboards (which is actually Business Intelligence and falls underthe Analytics umbrella). The fact is, it is both and so much more. Thedictionary defines analytics as “the discovery, interpretation, andcommunication of meaningful patterns in data.” This definition only identifies the computational part of the practice. To be leveraged successfully, analytics needsto be more than just the math and statistics. The analysis must combine judgment, common sense, and businessunderstanding in order to gain real meaning from data and truly inform thebusiness.
USE CASE: HOW PREDICTIVEMODELS IDENTIFY WHERE YOUR BUCKET LEAKS
Businesses will always have theories about why customersbehave in certain ways—like understanding why customers stop purchasing from them. It is natural for most to think that there isonly one reason a customer left when in reality, it is likely a combination of several. It is also common and, unfortunately, veryunproductive to have a hunch about why customers are leaving and then askanalysts to perform data mining tasks until the information needed to support thetheory is uncovered. This scenario is knownas a “fishing expedition,” and Scott Adams finds humor in this common behavior.
It is a challenge to identify all of the reasons why acustomer stops purchasing, especially with one-off bivariate analyses. Predictivemodels can help in these situations as they are meant to:
· Identify drivers of a specific customer behavior
· Predict future customer behavior
· Monitor customer performance
EXAMPLE:
A former client believed that price was the mainreason their customers stopped purchasing a certain product. So, they hired apricing expert to test various pricing scenarios and recommend which wouldretain their customers while maintaining revenue and profits. Unfortunately, the analysis and testing wereinconclusive. As they continued to see their customers leave, this client becamedesperate for answers. They engaged an experienced Analytics professional whodeveloped a predictive model on all available data sources. This modeluncovered multiple reasons why customers were leaving, none of which hadanything to do with price. Some of the reasonsincluded solicitation channel and payment method, things the client could quicklychange and correct. The client was able to immediately adjust theirsolicitation channels to those that acquired customers preferred and encouragedpayment methods that made it easy for customers to renew their orders. Themodel was also used to predict when a customer was likely to stop purchasing,so corrective action could be taken before the customer left. These improvementsquickly and significantly increased customer retention rates.
MOVING FORWARD
Leveraging analytics to retain your customers is just one ofmany use cases that can help a company thrive. Other pertinent topics to come include but are not limited to:
· Leveraging models to prioritize and invest inthe right strategic initiatives.
· Why and how you measure matters—defining KPIsthat accurately measure company performance.
· Calculating and predicting customer lifetimevalue post-COVID as inputs have changed.
· Understanding and leveraging behavioral andattitudinal segmentation.
· Integrating new sources digital data sources toimprove personalization.
Whatever your questions are, they can be answered with thedata you have and some analysis.
