Why can’t the big boys innovate?

This column attempts to tackle the lack of innovation in large corporations. There are many ways to approach this problem — using your left brain is one way.

We believe we can use deductive reasoning to achieve innovation. This only works for engineering; disruptive innovation is different. But most innovation is based on human intuition — the gut.

I recommend that you see the movie Moneyball. I’ve talked about this before in my column, but it’s worth another mention. The short narrative version is that you hire only baseball players who can get on base — not those with high baseball scores, high-priced players or the coach’s nephew. Seldom do all-star teams win. But if the batting average were 300, you can pretty much guarantee that at five times at bat, the player would get on base. That’s all you have to do to win at the end of the day. In the true story, the team won 20 games in a row using low-cost, “on-base” players.

In venture capital or with angels, we apply the same principles. Angels, by the way, are people who spend their own money in small sums to get a company started. The Breakfast Club, a small group investing their own money of which I was a part of, utilized the moneyball principle. Our batting average was about 300. This means that we had successes in one-third of the companies we invested in. We had to invest in five companies, each of which had a batting average of 300, to assure success in this statistical run. The only difficult thing with this situation is that if you have a second one that looks good, you must not invest in it. The reader is encouraged to look up the three-door Monty Hall problem for more information on this theory. The Breakfast Club ran this way, on an intuitive basis, for four decades and over 1,000 business plans. There were enough winners for us.

Some examples of single, innovative inventions are the programmable logic controller (PLC) and building management using Andover controls. Many of you know the PLC story: On January 1, 1960, we embarked on the development of the programmable controller and, to our surprise, it took off. We did it mostly to get rid of the problem of continually redesigning and did not consider it an investment. Andover controls was the PLC for building management. We sold the language to the home builder and contractor to eliminate the problem of redesign. An unexpected result from the Modicon PLC was Modbus, the de facto automation communications system.

Those were successes, but many adventures were mistakes. We recently felt that indicator lights would be a nice thing to have on the factory floor — no go. We also decided to build large-scale super PLC hardware with easy-to-use software — another bomb. We attacked automated vehicles circa 1965 — too early. We built a rotating floppy disk for the military in 1967. We felt that it was unusable for the commercial market because of memory density and unconventional hardware. The floppy was not a bomb, but for us it was. You win some, you lose some.

Some of the lessons we learned were that if there are no bumps in the road, you’re not moving. By this we mean you must maintain a risk/reward function and allow 70 per cent of your new ideas to fail to have one succeed. Start with five and you have an over 90 per cent chance of having one — but only one — winner out of the five. It takes time. On average, we like to have cash come back to the investors after seven years. We wish for five years, but it seldom happens. Another blocker is what Scientific American calls an “einstellung” — the German word for attitude. The article subscribes to the issue of good being the enemy of better. Conventional wisdom doesn’t necessarily win.

I’ll share two examples. I work in a barn with a lot of fluorescent ceiling lights. One of the lights in the far unused corner was blinking, so I decided to fix it — I turned the light off.

Another example concerns chess. One of my children played chess reasonably well. A good friend of mine is an excellent chess player. He bet that he could beat my kid five times in a row and if not, we would get a bottle of vodka. We kept getting bottles of vodka. This is how the kid worked it: He played the player, not the game. Using unusual pieces and annoying music helped. He also attempted to remove the opponent’s queen at all costs, especially if he was making the first move. The first move determines the game. Once he removed the queen, the game they were playing was not in the chess books. That means they’re playing each other. Think about that.

I also found that technology is irrelevant — what counts is the market and the value to the buyer. It’s annoying to have your operating system continually updated when all it does is make things more complicated for the user. Today, you can go to any modern group of geeks, describe the problem and they will build you a solution. Recognition in the marketplace is the key, along with statistical analysis. Most big companies have a cycle that goes something like this: seek an idea, present it to your boss, the boss can either say yes or no. If he says no, the session is over. If he says yes, it really means a request to his supervisor for a no. Even if it had a 50-50 chance, you will get a no. Any risk that doesn’t get a quick reward penalizes the innovator.

But let’s get back to the initial problem presented — large companies have trouble innovating. Being an innovator in a large company usually doesn’t work. Why? They want to solve a problem they think they have. That’s called engineering. Innovation is when you make a problem for the competition. It requires a revolution in thinking and understanding that you have a groove in your brain that says the old ways always work, but how well do they work and are they solving the real problem? Those are important questions.

So here’s my advice: start from the bottom; only do five projects; hire smart people; do not micro-manage; take your time; only one can win; repeat next year; and call me in the morning.

This column originally appeared in the May 2014 issue of Manufacturing AUTOMATION.