I'm Lovin' It

I earned my very first paycheck from my first real job at McDonald's. $3.03, net of taxes. I learned a lot in those nearly two years of working side by side with Ronald McDonald and the Hamburglar. I recall a certain amount of curiosity at how much measuring was taking place. I'm not referring to the amount of lettuce on a McDLT or the number of ice cubes in a large Sprite. During some shifts, it seemed that all some managers did was fill out reports -- product usage, product waste, drive-thru times, etc. I did not envy whoever got the mundane task of entering those handwritten reports.  

I remember how much our managers were rated on all those metrics. For example, when working the window of drive-thru, I made sure to push the order complete button on the monitor as soon as the order was in the bag, regardless of whether the car was actually at our window yet. I also remember that product waste had to be thrown away, not consumed by ravenous teenage employees. With one notable exception involving nearly a full batch of Breakfast Burritos (now Sausage Burritos), we stuck to that rule. 

All this is a lead-in to how I've also long believed that McDonald's actually acts on all those metrics. Sure, you can point to product misses such as the McLean Deluxe, the Arch Deluxe, etc. (Anyone try the Hula Burger?) But, they've also had some remarkable successes. In recent years, many of those successes have been on the operations side of things. What I love about this approach is that they're not just focusing on what people eat. Business analytics focused on operations includes the entire customer experience.

I go to McDonald's frequently. Not daily, but often. Let's just say that I used to have exact change for my usual order before I even arrived. And I bet 90 - 95% of my visits are via the drive-thru. Yes, I want the food to taste like my last trip, but I also measure the success of my visit by factors such as the following: (1) Can I understand the person taking my order? (2) Can I see my order on the screen to verify accuracy? (3) Do I have to wait an amount of time I arbitrarily consider "too long" to move through the line? Each of those factors plays into how I feel when I pull away from the golden arches, and none of them involves the actual food.

The JMP Blog from SAS has posted videos of Michael Cramer from McDonald's describing their operational research. I think the videos are spot on and succinct, and he doesn't reference SAS or JMP at all. Here are the links: 

What I love about these videos is that he successfully introduces analytical principles in just a few minutes. For example, he quickly shows the fallacy of just looking at averages? How many internal reports do businesses create that involve only averages? Sure, you'll pick up a few trouble spots, but you'll probably also waste some time on false positives. I recall recently analyzing our ETL logs to track down a suspected bottleneck. If we had attempted to just look at averages, we probably would not have so easily isolated the issue. Instead, we quickly found the bottleneck by analyzing the logs using distributions and including additional factors. 

As for the customer experience, what aspects of our environments could be designed differently? Not at the technical, data layer, but at the user experience level. Yes, we're delivering our product, but how could we deliver that product better, more efficiently?

Another huge benefit of metrics-based decisions to influence operational changes is that they allow McDonald's to be more profitable without raising prices. The increase in efficiency also allows capital initiatives to be self-funding. Think of the dual-lane drive-thrus introduced over the past few years (McDonald's is obviously not the only one to employ this method). The increase in throughput alone may have paid for the changes at each location. 

Using analytics effectively isn't just about the final product, whether you're selling food, widgets or services. It's about the full customer experience.

And, just in case you're wondering, in my neck of the woods, 2 sausage burritos and a large Diet Coke will set you back a mere $3.30 — regardless of which drive-thru lane you use.