Today's WSJ has an article about how Wal-Mart is moving towards computerized scheduling that micromanages schedules to match employee shifts to shopper volumes.
Early this year, Wal-Mart Stores Inc., using a new computerized scheduling system, will start moving many of its 1.3 million workers from predictable shifts to a system based on the number of customers in stores at any given time. The move promises greater productivity and customer satisfaction for the huge retailer but could be a major headache for employees.
The system… tracks individual store sales, transactions, units sold and customer traffic in 15-minute increments over seven weeks, and compares data to the prior year's, before scheduling workers.
Is this “Lean?”
I'm a bit torn on this one. From a lean standpoint, this sounds OK in one dimension — with lean, we always talk about matching capacity to demand, of matching the workforce to workloads. We don't want to pay people to stand idle, if we can help it. That's not revolutionary. The ideal, for a manufacturing environment, would be to level load production activity so employees can work consistent predictable shifts during the day, the month, the quarter, and the year.
But, some environments are tough to level load. Take hospital labs, for example. Hospital labs are notoriously busy from 4 to 7 AM (due to MD rounding patterns)… it's not practical to level load completely, so you have to match up your schedules to the busy times, to make sure you have more people on hand when there's more work to be done. The downside of the non-level workload is that you have slow times, which leads to wasted time and waste labor costs. In labs, Techs work fairly predictable (or at least planned out) 8, 10, or 12 hour shifts… so they have some slack time during their day.
The LAST thing I would do is to tell a lab to start jerking around employee schedules in an effort to cut costs “at all costs”. In this field, it's supply and demand, where, for one, no Med Tech is going to accept short shifts or weird short shifts. They'll leave and find another job if you ask them to work just 4 AM to 8 AM, then go home and come back from 2 PM to 6 PM. Not going to happen. Secondly, it's somewhat disrespectful to people to treat them this way, to say “we don't have direct labor work for you right now, so you're of zero value to us.” What I recommend to labs is that they use the slow times for problem solving, kaizen, training, etc. Labs treat their Techs as “fixed cost” pretty much and can invest in them as people and employees.
So What is Wal-Mart Doing?
So, that said, Wal-Mart is NOT taking that path (and that's probably not a surprise to most Wal-Mart watchers). Wal-Mart puts this in terms that sound customer friendly — making sure they have more staffing when more customers are in the store. As a customer, I would appreciate that, since it's frustrating to have long register lines, for example.
The article says:
But while the new systems are expected to benefit both retailers and customers, some experts say they can saddle workers with unpredictable schedules. In some cases, they may be asked to be “on call” to meet customer surges, or sent home because of a lull, resulting in less pay. The new systems also alert managers when a worker is approaching full-time status or overtime, which would require higher wages and benefits, so they can scale back that person's schedule.
That means workers may not know when or if they will need a babysitter or whether they will work enough hours to pay that month's bills. Rather than work three eight-hour days, someone might now be plugged into six four-hour days, mornings one week and evenings the next.
Some analysts say the new systems will result in more irregular part-time work. “The whole point is workers were a fixed cost, now they're a variable cost. Is it good for workers? Probably not,” says Kenneth Dalto, a management consultant in Farmington Hills, Mich.
The article continues with more examples of how this haphazard scheduling is hard on employees and how employees are unhappy. I don't blame them.
What Should Wal-Mart Do?
If Wal-Mart asked for my advice, I'd say there has to be some middle ground. Sure, the goal of properly staffing for customer loads is great. But, also have to respect your employees and maintain their satisfaction. The last thing you want in retail is grumpy employees (see Home Depot, but their employees are probably happy today with their CEO resigning… a story for another post).
Do your best to staff appropriately, but accept there will be some “downtime.” Use that downtime for training, kaizen, and other productive activities (maybe like stocking shelves) that can be done at any time. You won't antagonize your employees as much. Sure, you might have higher costs, but there's got to be some pay off from happier, more productive employees, who are working to help improve your business.
I guess Wal-Mart isn't like a lab. They can afford to treat their workers like interchangable retail-bots…. but they shouldn't, not if they were thinking lean. I wonder how Tesco does it, a great “lean” retailer from the UK?
Another Question: What about the Forecasts?
I'd also wonder if last year's shopping patterns are really a great predictor of this year's? What if there are different sales and promotions going on? The system might “work” wonderfully, but what if it's “garbage in / garbage out” like some supply chain scheduling systems? This Wal-Mart system is basically using a forecast…. is the forecast accurate? That's the bane of manufacturing systems and their fancy algorithms… inaccurate forecasts.
Here is Kevin Meyer's post on Evolving Excellence… same conclusion. I thought of him and his phrase about worshipping the almighty algorithm.
Gemba Pantarei also has an excellent post with their similar take on this topic.
It's good to see how all of us can have a slightly different take on the same article, I think.
Don't want to miss a post or podcast? Subscribe to get notified about posts via email daily or weekly.