Second verse, same as the first

The names are different, but revisiting a column from 1995 shows how little some things have changed for the companies that dominate the tech industry.

Esther Dyson The Observer
Esther Dyson is a former journalist and Wall Street analyst. Her Release 1.0 newsletter and "Release 2.0" book were read widely throughout the tech industry.
Esther Dyson
5 min read

In 2004, about halfway to the anniversary we celebrate here, CNET acquired my company, with its Release 1.0 newsletter and its PC Forum conference. I decided to look back to 1995 to see what I was writing about then. Here's an excerpted version of one column.

The takeaway today: The players have changed, but the issues have not. Turns out I was wrong about IBM. It went through a "government" period, but under new management, IBM is going through a rebirth as a software company partnering with, amazingly, Apple (among others). As for Microsoft, it's still big and important, but it needs to recapture the character of a startup after missing out on the mobile revolution that Apple and Google have so successfully exploited.

Microsoft: The government is us?

By Esther Dyson, 1995

Microsoft's biggest challenge over the next few years is not how to fight the government, but how to fight the forces of maturity that would turn the company itself into a quasi-government. There is a general creeping malady that causes most large organizations to resemble each other, assuming characteristics most identified with central governments: one set of products fits all customers, everyone must be served, stability is preferred over innovation.

In other words, the challenge is character, not strategy.

Overall, Microsoft has done a good job continuing to behave like a competitive company. It still has the paranoia of a startup -- or a dissident -- rather than the complacency of a monopoly or an entrenched government. Microsoft stands in marked contrast to, say, IBM. But the challenge grows greater as the company grows larger -- and as Bill Gates is treated as a visiting head of state wherever he ventures outside the US.

The salvation may be the Internet, in the same way that the Justice Department revived AT&T, but left IBM to wither of creeping governmentitis. (And now the Internet and telecom deregulation may spur AT&T to even greater vigor.)

While it may feel unpleasant in the short run, companies need continuing challenge to stay fit and grow healthily. Otherwise they turn into monopolies (we mean this figuratively; the legal definition isn't the issue here).

A monopoly resembles a central government, democratically elected perhaps (i.e. customers chose its better product), but with no system in place for new elections. It generally stops innovating, and tries to conserve its advantages rather than explore new markets. What happens next (sometimes much later) is usually a revolution.

So let's look at some other characteristics of central government -- and the contrasting behavior of Microsoft and IBM in each aspect.

First, a government attempts to serve all the people. It feels (properly) a duty to serve those people not adequately served by the market. IBM, for one, spends a considerable portion of its resources to serve customers it acquired years ago who are no longer profitable. It is still maintaining products such as IMS and OfficeVision, and is still working on OS/2. IBM feels a greater loyalty to its constituents (benefit recipients) than to its shareholders (taxpayers). If a particular program is not profitable, it can make it up from other product lines in the recentralized model the company is readopting.

Second, a government doesn't like to abandon mistakes; it tries to justify them. Like a government continuing to pour money into a failing aid program, IBM is convinced that one day its products will win. Besides, it has made a promise.

Microsoft, meanwhile, launches and retires products more scrupulously. It will push something for a few years, but it doesn't (usually) throw money down sinkholes. Unlike a government, it's innovative rather than predictable. It is happy to see new products wipe out old ones, and so it keeps beating competitors as well as cannibalizing its own products. Microsoft is still going after the best markets, rather than feeling an obligation to serve every niche. It wants to dominate the most lucrative markets, not all of them.


On the other hand, Microsoft does slip from time to time. Consider, for example, the DOS tax. DOS was something a majority of customers voted for, freely, over the years. Now it is virtually a "government" mandate: Until the Justice Department stopped the practice, Microsoft wanted to charge everyone for it (through an indirect tax collected by OEMs), regardless of whether an individual (or reseller) cared to take advantage of it. Another government-like move. Wisely discerning that the support burden for Microsoft Network might overwhelm it, the company decided to reduce demand -- not by pricing, but by rationing: The first 500,000 customers get the service; the rest have to wait.

And now, the Internet

The Internet gives Microsoft the challenge of an environment where monopoly is hard (impossible) to impose. But it also gives the company the opportunity to fight it out firm to firm, creator to creator, but in a world that values diversity rather than standards, variety rather than commodity.

Thus, although the Internet presents Microsoft with a huge challenge, it's also an opportunity to continue its growth in a new market rather than eke out the last profits from an existing one. DOS will continue to generate revenues -- as will the various implementations of Windows -- but the future is on the Net. And on the Net, Microsoft is not the incumbent.

When AOL complains that Microsoft can market MSN for free -- piggybacking on Windows 95 -- we wonder what would happen if Microsoft were to give Windows 95 away for free. AOL would have far more cause to object -- and so of course would the thousands of retailers and other resellers who would lose their cut. (It is after all better to compete with a P&L-oriented company than with a government.) Of course, in the long run, Microsoft probably will adopt the Netscape distribution model -- as will most software/service vendors.

While Microsoft continues to form new businesses and supersede the old ones, IBM is recentralizing. Bigness is good, proclaims Gerstner. We can offer worldwide solutions. That's true. But more and more of IBM's customers are downsizing, partitioning, fragmenting, working on local value. Microsoft seems to keep fissioning into new areas in which it can grow as rapidly as a small company.

The character issue

In the end, the issue is character. Can Microsoft persuade the world that it is entitled to govern -- even in a portion of a decentralized world? Can it fend off governments and regulators? Can it persuade application and content providers that it means them no harm? Can it learn to coexist in a fragmented world?

For both governments and companies, the new world of the Net means a different balance of power. In most fields, the majority will not rule; voters and customers will get their individual choices. Majority choices won't benefit from economies of scale the way they do in a world of standards.

Photo courtesy of Getty