Across from the Black Bull Inn, just off the Canterbury Road in Folkestone England, I used to get Maryland fried chicken from a Chinese take-out at a post office run by an Indian family. The chicken was notable mostly for the battered, deep fried bananas that were thrown in – looking like giant fried worms when first bitten by the uninitiated. The chicken was passable – mediocre by my taste but then I eat a lot of fried chicken. If that family is still running the post office, then their chicken is probably still being served to this day. After all, the chicken was not so bad that they would fail outright, and in any case they could survive as a post office or by selling Chinese food; yet neither did they delight their customers. Like so many businesses, they survived somewhere in the middle.
The problem is simple: We don’t get excellence because we don’t invite failure.
Let me explain.
If we seek out and eliminate failing projects, then we are left with only the clearly successful ones. And if we are then left with nothing, we have learned the valuable lesson that it is time to change direction. The more stringent our definition of success, the more failures we eliminate – and the better are those few projects that survive.
Thriving markets sometimes kill off failures for us. Highly competitive markets have high failure rates and therefore stronger survivors. It is no accident that Silicon Valley has good software companies; look at the high failure rates there! Back to fried chicken, you’ll find some amazing choices in Atlanta (and Seoul). A mediocre fried chicken shack in Georgia has the life expectancy of a balloon in the pokeweed. Thriving markets kill mediocrity.
Some companies harness the power of the market. They design stand-alone, specialized products and services, and so their failures are evident. For instance, consider the evolution of the file server - computers specialized to move data. When they came along in the early 1990s, some of the first innovators in this space were worried about competing with big, general-purpose computer servers made back then by Sun, DEC, IBM, and the like. In fact, some specialized file servers failed in that competition, and the companies that made them, like Auspex, are gone. This is the great strength of specializing: Specialists know when they are not competitive.
On the upside, the successful specialist is clearly best-of-breed. In the case of file servers, we saw this result for firms like NetApp and EMC. By making specialized file servers, these companies could easily judge where they did and did not have product-market fit. They then easily out-competed the general-purpose servers of the big players. Those big servers had their place, of course, at the heart of large distributed-computing systems and later as the backbones of the internet. But when it came simply to moving data, the big servers were outcompeted by the specialized servers.
Unfortunately, wrong-headed business strategy can prevent the market from eliminating mediocrity. This happens when leaders try to improve a product by adding more features. Think late-night advertisements for gadgets: “wait, and there is more!” sings the announcer, as the contraption is revealed to have yet another benefit. We end up with an “integrated” product that provides many different benefits to many different customers. Such products allow the losing parts of the overall product to stay alive when they should be allowed to fail. Integrated products allow features to subsidize each other. You are hoping that somehow the full set of integrated features will make up for the lack of individually-excellent features. Hermes’ cartoon of the one-man-band comes to mind. Amazing in some ways, the one-man-band never really caught on. One prefers a full band of individually best-of-breed performers. Best of breed trumps integrated.
So what about cherry picking from the world of innovators, and building a “suite” of best-of-breed features? This road to the integrated product is traveled by big firms acquiring up-and-coming specialists that have survived the market test. This strategy can work – as long as the business acquires best-of-breed products, and as long as the value of these products is improved by being in the big firm. We’ve seen this outcome at big technology firms like Cisco, Microsoft, Google, and Facebook. These companies acquire rising stars in the technology world, and then point to their growing portfolio as evidence that they are innovative. (Somebody was innovative, and their reward was to be acquired by somebody who then claims to be innovative.) Integrated products built this way can be effective, if you are well-funded and can spot the best-of-breed early on.
A cautionary note: You may think you and your firm can make an integrated product with many features, and all of them can be best-of-breed. You may hope for that outcome, but here’s your problem: The way you get best-of-breed is by discovering product-market fit, and such discovery requires a market test. But the market test is prevented when there are too many benefits to your product. You'll end up mediocre, because you failed to invite failure.
For some of the research behind these ideas, take a look at the work of Olav Sorenson.