That’s why, a couple of years ago, I got a call from The Coca-Cola Company (TCCC). A great torrent of Coke flows through McDonald’s every day. So TCCC was particularly concerned by a new finding: that consumers order a smaller size of Coke when passing through the drive-through than when ordering indoors at the counter. Multiply this difference (even if it’s just 3 ounces) by millions of drinks per day over thousands of outlets, and you get the idea.
All the obvious methods had failed. TCCC had collected every kind of quantitative data and subjected it to every statistical manipulation. Focus groups had been held. Consumers panels had been convened and consulted. The experts had been consulted.
All the obvious answers had been exhausted. No, it wasn’t that people were worried about spilling their Coke in the car. No, it wasn’t that they were concerned about bladders filling and perhaps exploding on route. This is was for both McDonald’s and TCCC a WTF moment.
So TCCC asked me if I could solve this problem? I said yes, because I always say yes. But I figured that if all other methods had failed, no one would be inclined to crucify the method (or me) if ethnography did too.
Off I went to Southern California. I found a McDonald’s with the right mix of consumers and a robust drive through line. The manager was a little mystified but entirely helpful. I decided to see if I could interview people in the drive-through. And predictably, this made people in the drive through line a little uncomfortable.
I was wearing a suit and tie, to make myself look more professional. Actually, wearing a suit in a drive-through line really just makes you look Martian. But that’s good too. To magnify the effect, I wore a little blue McDonald’s cap. Now, I looked like a complete asshole. To be fair.
Very shortly, I was feeling like an asshole, too. The longest you can talk to someone who is driving through a drive-through line is about 40 seconds. You have no longer introduced and explained yourself than the driver is ready to move on. You can walk along beside the car as if shuffles forward, but this is tricky because you have to walk backwards, and the lane is narrow.
So most of your interviews last about 22 seconds. It is hot, very hot. It is so hot that I am soaked right through and my camera person, Suz, has discovered that the sun block has run off her forehead into her eyes and she can barely see. At this point, we aren’t getting very much more on camera than we are in the interviews.
I can’t tell you the outcome of the research. That was bought and paid for by TCCC, and what are the chances that I would ever work again if I blabbed (and blogged) my research results. But, at the risk of blowing my own horn, I can tell you this. The idea finally came to me, out there in the broiling sun, and when I gave it to my client over the phone, she gasped. In my experience, this is the best sound a client can make. I have only heard it once.
But let me end on a methodological point. Ethnography must always start with an act of humility. You are talking to the respondent because they know and you don’t. They are the expert and you are the supplicant. You have to communicate this early on, especially if you are talking to someone who might be intimated by you or your appearance. Nothing quite gets this job done like a suit, tie and a blue McDonald’s hat. In these circumstances, I might as well have been dressed up like Ronald McDonald.
But there is a second grounds for humility, one that exists for all researchers. Every time I drive by a McDonald’s, I look to see if they embraced my strategy or adopted any of my recommendations. Five years later, they haven’t done a thing.