Everybody enjoys reading predictions. They are popular before the start of any new year. Some savvy prognosticator tells us which food flavors will be popular, which retail formats will emerge, which tech solutions will catch on, and so on.
Of course, grocery retailers can also make predictions by deploying predictive analytics throughout the organization. This practice mines data for information to predict trends and patterns of shopper behavior.
But the use of predictive analytics in grocery is modest to moderate at best. Some grocers make several mistakes when starting.
When researching this topic, experts told me dozens of mistakes that grocers should avoid. Some of them came up over and over. Here are the typical mistakes that should be avoided:
Not Developing an Adoption Plan: Grocers must identify the parts of the business to be addressed, list the steps to be taken, and define the value to be gained.
Not Getting Approval from Senior Management: To develop an analytic culture, the top executives must buy into the program from the beginning. Getting commitment to support its use throughout the organizations can make the difference between success and failure.
Not Having Required Resources In-House: Sophisticated analytical work calls for certain skill sets. Predictive analytics is not the core business of grocers. Bring in the experts for consultation or hire experienced people.
Not Giving It Enough Time: Developing predictive analytics and integrating it throughout the organization takes time. Don’t be impatient and expect results too soon. It takes time to mine the data and develop actionable insights that make a difference.
Not Using Insights Correctly: Grocers need to be wary of broad insights that don’t apply to a grocer’s shoppers, or launching initiatives that would not be supported by most of them. A national trend doesn’t always manifest itself locally.
Now that we know what not to do, here’s a description of predictive analytics from Sanjay Kupae, senior product manager from Manthan, an analytics consultancy:
“Successful analytic deployments are owned by marketers or merchandisers, articulated in business terms, and treated as a core capability of their every-day functions, and embedded within their processes as systems. Analytics roll-outs are layered in a manner that campaign managers and category planners interpret data through friendly and engaging dashboards that surface the underlying algorithms, while the analysts and data scientists are able create, modify and fine-tune models without encumbering business users.”