Big Blues: Why IBM Is in Trouble

By Robert X. Cringely

http://www.pbs.org/cringely/pulpit/pulpit20060518.html

No word this week from Sam Palmisano. The CEO of IBM was strangely silent following my column last week about Google and Microsoft that also touched on the malaise at IBM -- a malaise very much of Mr. Palmisano's making. But the troops inside IBM Global Services DID reply, and uniformly supported my grim news from last week that their company has entered a death spiral of under-bidding and then under-delivering.

Why would a company DO that? Why would they compromise a reputation built over decades? Because decades no longer matter to publicly traded American companies. All that really matter are fiscal quarters.

Much of this began on Lou Gerstner's watch. And while I am generally a fan of Gerstner, who retired from IBM at the end of 2002, Sam Palmisano is generally reaping what Gerstner sowed.

Lou Gerstner was the first-ever outsider brought in to run IBM, and did so at a very dark time in the company's history; John Akers had self-destructed while trying to carry out the vision of HIS predecessor, John Opel. Are we beginning to see a trend here?

When Opel was CEO and IBM was almost strictly a mainframe computer maker, the company went through a huge growth spurt based almost entirely on what the first President Bush would have called "voodoo economics." IBM began urging its largest customers to switch from leasing their mainframes to buying them outright, which overnight raised revenue from those customers by ten or more times. IBM had a huge sales surge, generating crash that made the company look richer than ever, but also had to be put to use in the business. What made this voodoo was the company's expectation that this sales level would be somehow sustainable even after all the leases had been bought out. By that point, of course, Opel was gone, Akers was in, and IBM was better known as a PC company -- a PC company that couldn't make a profit.

Akers was clueless to fix the problem, but Gerstner wasn't. He downsized the company by more than 100,000 workers, spun off several business units, and most significantly, built the very services business that is now causing such headaches.

The attraction of services to IBM was based on the inherently high profit margins of that sort of activity. DOING things turns out to be a lot more profitable than MAKING things, and a LOT more profitable than inventing things.

IBM's gross profit margin is around 36 percent, which means in the very simplest terms that for every three dollars the company takes in, one dollar in profits is generated. Inventing things makes IBM a lot less money than that. One of the things Gerstner did when he took over the company was he started aggressively licensing IBM's patent portfolio, taking it from generating about $30 million in revenue per year to over $1.5 billion, which was touted as nearly 100 percent profit. The only expense was for the legal department.

But inventing things costs money -- about $5 billion per year for basic research at IBM. IBM saw this as simply a cost of doing business for an industry leader; hence it's being ignored in the patent revenue calculation. But the money was still being spent and had to come from somewhere, so this technology licensing business was generating $1.5 billion in profit for $5 billion in expenses for a gross profit margin of around 30 percent. That means that for IBM inventing things -- while still seen as vital to the identity of the corporation -- was actually a drag on earnings. The same was true (but even more so) for almost all of IBM's hardware businesses.

Once you have eliminated as many of these business lines as you can without compromising the very identity of the company, the only way left to give Wall Street the quarter-on-quarter earnings growth it likes is by taking on more high margin service business, which is exactly what IBM did, first under Gerstner and then under Palmisano.

This is not inherently bad, but if you combine it with a lack of vision, the strategy eventually falls apart. Remember that IBM's mantra in this was to "fulfill customer requirements." That's a clever slogan that sure sounds like it is based in six-sigma customer service, but it really isn't. That's because most customers don't really know what they want. And if you count on fulfilling these unknown and certainly un-vetted requirements as your corporate raison d'etre, well, it leads to where the company is today -- milking (and ultimately bilking) its customers.

Here's how one IBM employee put it recently as he resigned: "Unfortunately, I see IBM as a place run by salespeople and project managers with a sell and install mentality, even in services. There is no technical leadership, innovation or proper understanding of our customer's needs and requirements. The architecture profession is dysfunctional and cannot remediate itself. These factors may change, but not in the short term and when it does it is likely to be brutal, and I'm not patient enough to wait around until it does."

And here's word from another IBMer who is sticking around for now: "As part of the cost cutting, IBM cut some of its billable rates to help get business. The problem is, consultants are expected to bill more and more hours to make up for the lower rates. For some (including myself), we're expected to bill 93.5% of 2080 hours yearly. The problem is, when you factor in vacation, a sick day or two, etc, that leaves almost no spare time. What's happening is that IBM is no longer providing training to their consultants, expecting us to pick things up on our own. This is leading to a much lower quality of work on projects. In addition, the demand for more billable hours is resulting in some ethically questionable actions. Starting four weeks before the end of each quarter, we start get emails asking us to try and bill more hours. They always include statements saying "to help improve customer satisfaction and meet deadlines", but with a wink-wink it's implied that you add on an extra hour or two. The resulting billable hour crunch has also led to less people exceeding or meeting their goals, leading to an overall lower yearly bonus (called variable pay). Many people are quitting IBM, and IBM is now in a hiring crunch because it can't fill projects. The result is that they're stuffing anyone available onto projects (regardless of skill level), again lowering the quality of our deliverables. The lowering of bonuses and increased utilization has prompted many (former Price Waterhouse Coopers) people to jump ship. So IBM is sacrificing the long-term health of IBM Global Services, to keep up the quarter-to-quarter results. Delivery quality is down, employees aren't getting trained in newer technologies because of the crunch to get more billable hours, and people are leaving IBM because of the impact on pay and overall low morale."

There is a perfectly good solution to IBM's situation, but it requires candor, courage, and personal accountability on the part of management (that would be you, Sam). IBM first needs to be honest and realistic about its situation. IBM needs to put together a business improvement plan and invest in itself. IBM can go to Wall Street and say, "We've neglected some of our strategic core businesses. We are going to invest in them and bring them back to health. For the next "n" quarters we will be reporting lower profit numbers. This investment will position us to increase future revenue and deliver services at a lower cost, and generate more value to the shareholders."

Wall Street actually likes plans of this nature and -- when they are explained properly -- responds by not beating-up the stock too much. The price might even go up.

The trick is to be able to execute on the plan and deliver the goods. IBM is good at overpromising, overspending, and underdelivering on its proposals. They don't know how to design and implement small, super cost-effective projects that result in real benefits. They are probably incapable of fixing themselves.

For example, IBM's vaunted Global Technology Center (GTC) in Austin has created a new "help desk" system called eESM. It is big. It is expensive. It is gold plated. It is millions over budget. It is years late. It is unreliable. It doesn't provide the features the business needs. The intent was to put every customer on eESM, then move all the help desk and server admin functions outside of the U.S. (If you are an IBM USA-based server admin, watch out!) The problem, of course, is that eESM won't really lower costs at all, and probably won't make anyone happy. There is talk of scrapping it altogether.

Think about the auto industry. Why is it possible to ship iron ore and coal to Asia, where they can build a car, then ship it back to the USA and sell a better product at a competitive price? The answer is QUALITY. When your QUALITY is crap, your costs increase. You compensate by throwing more and more people at the problems that result from poor quality. When you try to cut costs, the problems increase and the quality suffers even more. IBM is in such a death spiral, and it is a direct result of a management decision to sacrifice quality and NOT to properly invest in improving and automating its services businesses.

IBM project management is not based on business results. It is based on documented deniability. A successful IBM project is completing everything as originally documented. If it works or not, it doesn't matter. The folks in the GTC will be quick to tell you eESM is a great success and it exceeded all the requirements. Sadly, its users needed something better and a lot less expensive. In most businesses, this project would be considered a failure. In IBM, it isn't. And the organization keeps spiraling towards the ground.