Most books are reviewed when they are printed, or shortly thereafter. But once in a great while, it is worthwhile to review books that were written long ago but are relevant today. There aren’t many books that fit this category; however, there is one seminal book about the computer industry that deserves as much attention today as it did thirty years ago. That book is The Mythical Man-Month by Fred Brooks.

IT managers need this book.
 
So what is so important or profound about The Mythical Man-Month?

Fred Brooks, it seems, was associated with some early IBM development efforts for software. If memory serves me correctly, the software project that Brooks was associated with was the development of the IBM 360 operating system. Prior to the IBM 360 operating system, each operating system had its own software. Each operating system was different – in some cases fundamentally different – from each other operating system. This made life more difficult for the systems programmer, but it wasn’t just the systems programmer whose life was difficult. The application programmer was the real person whose life was even more difficult. Each application had to be custom coded for each operating system. Think about it – no SAP. No Oracle. No spreadsheet. No industry-wide application for anything!

Most of us simply cannot imagine such a world.

But it was not the profundity of the operating systems software that made Brooks’ book the milestone book that it was/is. It was in The Mythical Man-Month that Brooks made has famous observation that when a software project got into trouble, adding people increased the trouble. If a software project was late, adding people to the project made the project even later.

This bit of wisdom flies in the face of what most managers do when a project is in trouble. In most cases, when a project is in trouble, managers add people; however, in the case of software, this is absolutely the wrong thing to do.

The world is indebted to Brooks for this observation.

And there is a perverse and unintended consequence of Brooks’ observation. Suppose a corporation has two software projects running – project A and project B. Project A is run by experienced, competent people who know what they are doing. Project A is on time, within budget and meets every milestone. Project B, on the other hand, is run by incompetent, inexperienced people. Project B is late, over budget, and never meets a milestone. So what does management do? Why, they give resources to project B. (They might even take resources from project A, in the hopes that this will get project B off the rocks.)

And at the end of the day, who gets the larger reward – the management of project A or project B? Why, the manager of project B gets the bigger reward because in our system, people are rewarded for how many people work for them.

Talk about giving rewards to the wrong people. Talk about systemically reinforcing the Peter Principle. The people who are competent and who do a good job are not rewarded, while the people who are incompetent and who do a terrible job are richly rewarded.

If you don’t have The Mythical Man-Month on your bookshelf, you ought to go out and get it.

Recent articles by Bill Inmon

• Predicting the Future
• Pet Peeves
• Wrestling with Corporate Contracts
• Managing the Unstructured Volume