AI for Retail

CHAPTER 18AI for Labor Budgets and Scheduling

Current Roles, Processes, and Inefficiencies

Most labor budgets today are based on the sales volume of that store. The more sales, the bigger the labor budget. This is flawed however. Let's say a store has a bad month or year because of a bad store manager or bad luck resulting in low OSA and thus lower sales. Lower sales lead to a lower labor budget. Lower labor budget results in even lower OSA since we have even less labor to fill the holes, and thus even lower sales. This is called the “Store death spiral.” So this strategy makes no sense in the long term. Labor scheduling should be bottom‐up. It should be based on the work needed to be done in the store and go up from there. If you have more outs, you have more work to do, so you should allocate the labor required to work those outs and for the store to get back on track, instead of starving the store of further labor hours.

Labor breakdown is typically done based on labor standards for specific tasks such as how long it takes to perform a checkout or an out‐of‐stock scan or how long it takes to restock a product. They are seldom right and infrequently looked at ever again. We have seen a strategy where operation leadership update labor standards in a spreadsheet, reducing them let's say by 10% with some new process they announced, just so they can cut hours and claim a win for the operations leadership team. If it used to take 4 hours to scan the entire store for out‐of‐stocks, ...