In case you hadn't noticed, many of the largest Internet backbone companies are in big financial trouble. Several, like Global Crossing, Williams Communications, and of course, Enron, are in bankruptcy, while stalwarts like MCI and Uunet are just limping along as part of a Worldcom that might, itself, be fatally injured. Conventional wisdom is a bit conflicted about what this means to you and me. On one hand, new owners will eventually come along and they'll have a lower cost structure, having bought their networks for pennies on the dollar. This could drive down the price of Internet access. On the other hand, these same new owners will want to make money from their network properties (something the previous owners never did), and that suggests higher prices. I think higher prices will prevail, in part because they already seem to be rising, and also because these bankruptcies are likely to lead to greater industry consolidation, which means less competition. But it goes further than that: As the Internet industry consolidates, the surviving players will not only want us to pay more money, they'll want us to rely on them alone for certain types of data services. It is already happening.

For years, we talked about how the Internet was growing so quickly, and how that meant bandwidth would get cheaper over time. Well, it did for awhile. But just as companies like AOL are starting to raise monthly rates, so too the broadband companies are looking to augment revenue. Some are raising rates while others are limiting services. In California where I live, SBC Communications is by far the dominant DSL provider, and SBC has found a third approach to this problem — they limit someone else's services.

Huh?

Here's the deal. If you buy broadband service in California or any of the many other U.S. states where SBC does business, the purchase often isn't directly from SBC but from a local ISP that buys DSL service from SBC. So 90 percent of the DSL circuits in California, for example, go through SBC, but many users of those circuits don't know that. In this arrangement, your local ISP is SBC's customer, and other than providing a local DSL circuit and maybe some Internet bandwidth, SBC is supposed to stay out of the ISP-Customer relationship. But now we have Project Pronto, SBC's grand network upgrade, with its Broadjump Service Navigator software. Service Navigator is a tool to help SBC manage its DSL lines, and one of its more interesting capabilities is to make it just as easy — sometimes even easier — to reach SBC's servers than your local ISP's. Service Navigator can be used to guarantee quality of service to SBC, often at the expense of the local ISP, who thinks the end-user is his customer, not SBC's.

I'm talking about SBC here, but this technique is not new and is practiced by other telcos, too. The point of it all is that when we start downloading commercial movies over the Net, Service Navigator will make it a lot easier to download them from SBC rather than from your local ISP or from some anonymous FTP server in Finland. If the Internet really does become a true entertainment medium for more than just looking at dirty pictures, companies like SBC want to be sure the download business is theirs.

The other side of this is limiting service. If you are a broadband user, look at your service agreement and see how much data you are allowed to download each month, and what it will cost to download more data than that. Then divide into 4.7 gigabytes the total allowable data. and tell me how many bootleg DVDs you'll be allowed to download each month. My calculations say the number is one, uno. This means two things: It means most DSL users won't have enough bandwidth to be major net providers to peer-to-peer video sharing services; and it means that companies like SBC that can use Service Navigator to take download bandwidth off the meter will have a killer advantage over both peer-to-peer and other download services.

The telcos will take these two concepts to the movie studios and sell a brighter vision of the future where maybe there aren't any movie theaters left, but Sony still makes lots of money. They'll eliminate the competition of peer-to-peer file sharing at the service level where users really have no power.

Part of this story is playing out right now in Canada, where Rogers Cable (cable modems) and Bell Canada (DSL) have just announced a tiered pricing schedule capping downloads at various levels from one to five gigabytes per month. If you want to download more than your allotment, the charge is $7.90 (Canadian) or about $5.00 (U.S.) per gigabyte, which means the second pirated DVD you download will cost $25 with none of that going to the movie studio. It is a brilliant plan to give us a fire hose, but then not let us drink.

Maybe Canada is a proving ground for this strategy. After all, SBC Communications does own 20 percent of Bell Canada.

Changing subjects, I wrote a column about problems at IBM just prior to the tragic death of my son Chase. That column produced a great response from inside Big Blue and out that I think warrants a short revisit to the topic. My point was that Sam Palmisano, the new IBM CEO, has his work cut out for him to bring real efficiency to IBM. I don't think he can do it.

I received a number of great e-mails on the state of IBM following that column. One writer had been recently laid off from IBM. He shared with me the stories of his now unemployed comrades and how they were treated by the company. Last year for example, many were asked to bill 2,010 hours of "productive" work. You need to do some math to really understand the implications of this directive. You can take two weeks off for vacation or illness. Any more time off has to be recovered by overtime. There will be no training. If you have more than two weeks of vacation, you have to put in even more overtime. Thousands of IBM employees were asked to work 45 to 50 hour weeks all year without any additional pay.

There are two ways one can hear this story. If you worked for decades at IBM, it is a story of horrible cruelty, of a good company gone bad. If, like me, you come from Silicon Valley — the real Silicon Valley, where we work hard for our Porsches — it sounds like a joke. Only two weeks of vacation per year? Let's see. My last real vacation was in July, 1989! Working 45 to 50 hours per week? I've been doing 80 hours per week for 25 years!

But real IBMers expect more and have traditionally received it. They accepted lower pay and frequent transfers as the cost of working — but not too hard — in a welfare state. Forced overtime, compromising family life, poor compensation, and poor severance packages are not the actions of a company that values people, and it is this change in IBM that some readers decried.

But many readers wrote me to defend how IBM is managed and runs its business. They generally had no adult work or business experience outside of IBM. They thought I was badly misinformed. Unfortunately, they had no idea there were better ways to do things and what they were defending really wasn't worth defending. To face change you must challenge your perceptions. You must be willing to question your methods and be open to different, even radical ideas. Many of my e-mails were from IBMers who were not ready to face the change challenge, which may be closer to the heart of the company's troubles than anything else.

Some readers told me the core of IBM's management problem is the need to have a large department with lots of people working for you. The size of your department is the most important management criteria in IBM. Few grasped the concept that every product and service needs to not only to be profitable, it needs to be more profitable each following year. The only way you can keep improving profits is to fully understand your costs and use that data to manage improvements. Sometimes managing improvements means investment and change. When a large part of your management organization and none of your non-management team have no access to your costs, you cannot really improve things. It is simply impossible. The only cost reduction tool you have then is to cut staff. It has become common practice in IBM to cut staff and the services they provide, and keep cutting until something breaks. While this makes the company look good for the short term, it is a great way to nuke your long-term business. When services are an extremely important part of your business, nuking your future is not a good strategy.

If you don't know your labor and financial systems are antiquated, if you don't know the value of providing high quality cost data to all decision makers, you can't even know how to make your business run better. IBM needs to question its management practices. It needs to change its perception of how to run a business.