clock menu more-arrow no yes

Filed under:


In the corporate world, when someone makes a mistake everyone runs for cover. At Microsoft, I try to put an end to that kind of thinking.

It's fine to celebrate success, but it is more important to heed the lessons of failure. How a company deals with mistakes suggests how well it will bring out the best ideas and talents of its people and how effectively it will respond to change.The message I want a manager to communicate is, "I don't blame anybody in particular for this problem. What I care about is how well we rally around to come up with a new approach to resolve it."

Back in 1984, after Microsoft released the version of a spreadsheet called Multiplan for the Apple Macintosh, we discovered it contained a bug that could damage data.

When members of the Multiplan team informed me of the problem, they asked if we should send a free corrected version to the product's 20,000 customers.

I said yes. There was no argument, no discussion. The answer was obvious.

Before he left, the team leader said, "But this is going to cost a lot of money." Indeed, it was - almost $250,000.

"Just because it's bad doesn't mean that there is any room for discussion," I replied. "One day you come in to work and lose a quarter of a million dollars. The next day you come in and you try to do better."

Apparently people expected a bigger reaction from me, but there was no value in a bigger reaction. There wasn't even value in spending more than a minute in the meeting.

Reacting calmly and constructively to a mistake is not the same as taking it lightly. Every employee must understand that management cares about mistakes and is on top of fixing problems. But setbacks are normal, especially among people and companies trying new things.

When employees know that mistakes won't lead to retribution, it creates an atmosphere in which people are willing to come up with ideas and suggest changes. This is important to a company's long-term success. And drawing lessons from mistakes reduces the possibility that errors will be repeated or compounded.

I've made some expensive mistakes myself. One also had to do with Multiplan - the very first version, which came out in 1981 and soon was running on more than 100 different computers, including the then-new IBM PC.

My blunder was to insist that Multiplan run on the Apple II, a popular but underpowered computer. To make it work, we left out some potential features that required more power. Lotus, a new company at the time, didn't make the same mistake. Lotus 1-2-3 was designed expressly for the more-powerful IBM PC. The Lotus spreadsheet ran faster than Multiplan and included simple charting and database features. My mistake - aiming Multiplan too low - knocked it out of the spreadsheet sweepstakes.

Recently I was involved in making the opposite mistake. Microsoft released a new version of Word for the Macintosh that required more machine resources than a lot of our users had. This set us scrambling to get out a modified version.

Fortunately Microsoft's board of directors isn't going to fire me for participating in that mistake. (At least, I don't think they are.) But not all chief executive officers can feel so secure anymore.

In recent years, many CEOs have been replaced as corporate boards worldwide have become more active, partly in response to shareholder pressure.

Recently, for example, Maurice Saatchi was ousted as chairman of Saatchi & Saatchi, Joseph Antonini was pushed out as president and CEO of Kmart Corp. and William J. Agee was fired as chairman, president and CEO of Morrison Knudsen Corp.

Although one can debate these particular board actions, in general it's probably good that CEOs are at risk. They get paid the most, and having the right person is vital. But even CEOs should be allowed some mistakes.

John Roach, the CEO of Tandy Corp., endured criticism a few years ago for lackluster performance at Radio Shack stores. I know John from the early days of the PC industry when we used to negotiate with each other. I was glad when his board members stuck by him - and I'm sure they're glad too, because under his leadership, Tandy has roared back with two new national chains of technology superstores, Computer City and Incredible Universe. Both are big successes, and John deserves the credit.

Frankly one of the challenges facing Microsoft is that many of its employees have not suffered much failure yet. Quite a few have never been involved with a project that didn't succeed. As a result, success may be taken for granted, which is dangerous.

With this in mind, we have deliberately recruited a few managers with experience in failing companies. For example, I hired Craig Mundie at the end of 1992 to oversee many of our efforts to develop the information highway. A decade earlier, Mundie co-founded Alliant Computer Systems, which eventually went out of business as the market for supercomputers dried up.

Mundie understands his mistakes and drew keen lessons from them. I think this is a reason he has proven to be among our ablest vice presidents. I'd hire 10 more just like him if I could.

When someone has run a money-losing operation, he or she has really had to ask, "OK, why do we have this many engineers? Who are we selling to? Who could we ally with?"

When you're failing, you're forced to be creative, to dig deep and think hard, night and day.

Every company needs people who have been through that. Every company needs people who have made mistakes - and then made the most of them.