Walk into a store run by a company that has spent the last decade digitizing everything. The shelf knows its own count. The price tag rewrites itself overnight. The marketing followed you in from your phone. Then look at the schedule that decides whether a single human is standing in the aisle to help you buy something. It came off a spreadsheet and a store manager’s instinct on a Thursday afternoon.
That is the tell. You have instrumented the entire store except the part that actually converts a shopper into a customer.
I spend my days in the forward-deployed seat, standing up production AI inside enterprise retail. And the pattern is consistent: the labor model is the last analog system in the store. Everyone knows it. Almost no one treats it like the priority it is.
The biggest lever is the least instrumented one
Deloitte put the quiet part in writing this summer. Labor remains the largest controllable operating expense for many retailers, yet it is the least modernized. Read that again. Your single biggest controllable cost is also the one you have left running on habit.
This is not because the technology does not exist. It is because retailers convinced themselves they already solved it. Over the past decade, retailers have materially improved how store labor is planned and scheduled. Standards-based scheduling and auto-generated schedules are now common among large retailers, enabling quicker, compliant scheduling while unlocking 0.5 to 2.5% labor cost optimization.
Half a percent to two and a half percent. Real money, but small ball. And here is the trap: most teams book that win and move on. Deloitte says exactly that. Leading retailers are beginning to layer in AI, using real-time task prioritization and labor insights to improve day-of execution. For most retailers, however, this is where the progress stops, and it may not be enough.
It is not enough because you are still framing labor as a cost to trim. That framing is the whole problem.
Scheduling is not overhead. It is where conversion gets decided.
Reframe it. The schedule is not an HR chore. It is a forecast, and it is the earliest point in the funnel where you decide whether Saturday at 2pm converts or leaks.
Your associates already see this. In a 2025 Logile survey, 77% of US retail store associates said their store regularly loses sales because of poor scheduling decisions. Three out of four people on your floor are watching customers walk out of understaffed aisles and abandoned checkout lines. They are telling you the labor model is broken and no dashboard is listening.
The academic work says the same thing in colder language. Researchers have documented for years that understaffing directly suppresses sales and profitability, and that traffic and coverage move conversion rates. The economics are settled. What is missing is the instrumentation and the will.
And the store has changed underneath the old model. Associates now fulfil online orders, handle complex inventory workflows, and deliver real-time customer service, all within the same shift. Scheduling alone cannot keep up, and the technology market is responding. You cannot schedule a 2026 store with a 2012 mental model of “put bodies on the floor.”
Run the CODN before you run the pilot
Here is where the Cost of Doing Nothing framework earns its keep. CODN is not a soft number. It is the compounding delta between the retailer who treats labor as a live forecasting problem and the one who treats it as a weekly formality.
Do the math your finance team will respect. Take conversion lost to understaffing at peak. Take shrink that climbs when the floor is thin. Take the turnover you drive when you schedule people into burnout because the model never saw it coming. None of those sit on the labor line. They hide in same-store sales, in shrink accruals, in recruiting costs. That is why they never get funded. And that is exactly the CODN: the loss you are already absorbing because it does not show up where you are looking.
The playbook is not exotic. Connect the workforce system to the signals you already trust. Invest in AI-enabled labour forecasting, automated scheduling and skills-based matching to improve scheduling accuracy and associate agility, and ensure the platform integrates with POS, IoT, and store operations systems. You have the traffic data. You have the sales data. You have the inventory tasks. The schedule should be a downstream product of all of it, not an upstream guess.
You have already proven you can instrument a store. You did it for the shelf, the tag, and the ad. The associate deserves the same rigor, because the associate is the only part of that list that actually closes the sale.
Digitize the last analog system. The retailers who do it in 2026 will not be talking about labor cost. They will be talking about conversion.