Nassim Taleb wrote a book named The Black Swan: The Impact of the Highly Improbable. As you can tell from the title, a Black Swan is what Nassim calls a highly improbable event. He asserts that these events are highly improbable, not predictable and they drive much of what goes on around us.
Many successful companies and products are the results of Black Swan events. As much as we’d like to think that there are formulaic ways of building successful companies, there isn’t. The best formula is lots of experimentation. The successful companies are the one experiment that works out of many. We just never really see the many because they die off before we ever take notice.
Many successful companies get their innovation wrong after they’ve become successful. They try to innovate from the top down. That is, high level management sit around at HQ thinking it’s their jobs try to dream up the next big thing. They forget where they started.
They create tests in a serial fashion and saddle them with arbitrary constraints like keeping the new product true to the brand, or sticking with their line of business because that’s what they know best or trying too early to make sure the product results can be reported through the company’s gargantuan enterprise accounting system, instead of just buying a copy of QuickBooks.
They forget that their initial success came from a long-shot experiment. I suggested that the failure rate is very high and the best thing companies can do is to allow for as many experiments to take place as possible and don’t worry too much about things like brand image and worry more about what customers think.
Tonight, I saw this in action at McDonald’s. CNBC had a show about the history and evolution of McDonalds and it reminded me of these thoughts on innovation.
McDonalds resulted from a Black Swan. Out of the thousands of burger joint experiments taking place around the country in the 40s and 50s, McDonalds experienced a couple of black swan events. First, the McDonald brothers tweaked their concept and made a nice, mulit-unit business. Second, it attracted the attention of Ray Kroc.
Even today, McDonald’s seems to intuitively understand the importance of many trials to find the next big things. While interviewing the current executive chef of McDonald’s, we were supposed to be impressed with his credentials. He graduated top of his class from a top culinary school and has prepared food for high restaurants like the Four Seasons.
That isn’t what impressed me though. What impressed me was that the chef seemed to understand that it wasn’t his culinary skills that led to successful recipes, rather it was the many trials. He said they test about 1,800 new recipes each year in his kitchen and only 3 or 4 make it to the restaurants. Of those fewer become successful.
Based on what I know of McDonald’s, I doubt that the experiments are limited to one test kitchen. I’m sure franchisees do some experimentation at the store level to find successful regional menu items and other items that might go big. I believe their McCafe coffee started as an experiment in their Australian stores, but I might wrong about that.
I wonder if McDonald’s recipe success is a proxy for the ratio of successful experiments to duds. I wonder if there were about 2,000 burger joints that eventually went away or continued operating in a small number of locations – experiments that may or may not have earned the owners a good living, but never took off on a wide scale – for every success story like McDonald’s, Wendy’s and Burger King?
This is all good stuff for company managers to think about. How is innovation managed at your company? If the company is constraining the number of experiments it runs, or dictating from up top how the experiments should be conducted, the company will find itself in a revenue plateau and then decline.
Breaking out of the plateau means opening the innovation spigot. That doesn’t mean recklessly spending money. It means changing the way managers think of innovation. Rather than thinking of it as a function performed by a group at HQ, maybe something everyone should think about.
Run lots of experiments, as cheaply as possible, and see how customers respond.
Realize that only a small fraction will show promise.
Don’t stop trying.
Don’t constrain your experiments.
Innovation doesn’t have to take place within your organization. Whether your company is experimenting much or not, experiments are taking place all the time. It never hurts to learn from those or acquire the experiments that show promise. McDonalds learned from Starbucks that people will pay more for a good cup of coffee and coffee-based drinks. They also bought into Chipotle early as that concept was showing promise.