More Evidence of a Maturing Software Industry

A front-page article in today's Wall Street Journal (free link should work for a week) headlined "Large Software Customers Refuse to Get With the Program" brings me back to an article I wrote a couple months ago about the maturing software industry . The thrust of the WSJ article is that (a) large enterprise software companies are having a hard time getting customers to upgrade, so they're (b) raising support costs and dropping support for older versions in a bid to raise more revenue, which is leading their customers to (c) switch vendors or go to internally supporting the software because they no longer see the value in upgrading. None of this is unexpected, and it is symptomatic of a seismic shift in the software industry which will be felt particularly hard in Redmond.

The software industry is maturing, and overall growth will be very hard to come by in the future (niche software markets can still grow, and some could become sizable). In a mature industry, customers tend to be focused on costs rather than benefits. Benefits are well-understood, and not unique from one vendor to another, so the customers want to get the benefits with the least cost. Compare this to an immature (growth) industry, where benefits can vary widely from vendor to vendor and version to version, and the benefit of a new installation or upgrade is much greater than when the industry is mature.

One of the consequences of the maturing software industry will be dramatically shrinking profit margins. In its most recent quarter, Microsoft had an operating margin of 38%, a ratio which is far above the more normal range of 10%-20% for a reasonably prosperous company. While the figure is not listed in Microsoft's SEC filings, I've heard that the operating margin for the core Office suite is in the 60%-70% range.

Microsoft has maintained its remarkably high profit margin because of the exceptional lack of competition and the ability to get customers to upgrade in the two markets which generate nearly all of Microsoft's profits: Office and Windows. Cracks are beginning to appear in both the competition and the upgrade front.

On the competition front, a shift in actual market share isn't required to force Microsoft to lower prices. All that is required is a credible enough alternative for large customers to hold Microsoft's feet to the fire. OpenOffice, MacOS X, and Linux are all becoming credible, and are capturing a few points of share. Microsoft may deride OpenOffice as being the functional equivalent of Office 97 (and they're right, though OpenOffice mercifully omits the paperclip), the simple fact is that Office 97 has enough functionality for the vast majority of users. And while Linux--even in its most friendly incarnations--is still too geeky for the average technophobe, OSX is perfect for luddites.

At the same time, Microsoft seems to be running out of "must have" features to incorporate into new versions. When the vast majority of requested features are already in the product (which begs the question of why are people still requesting them--but that's a different article); when the last Must Have version of Windows was Windows 95; when Microsoft has essentially halted development of Explorer; what's to induce the customer to upgrade? In fact, the opposite seems to be happening: Microsoft has been pushing back the next major release of Windows, aka Longhorn, at the same time it struggles to fix security holes in XP.

I serve on the Technology Committee of a local school, which runs a mixed OS environment (teachers on MacOS 9, administration on Windows, servers on Linux). We were discussing the question of whether to move to an all-Windows environment, and many of us were shocked when the network manager said, "I don't care if we run Windows or Mac, but if we switch to Windows, you'll need to budget for five more people in my department." As he explained, he was able to manage 50 Windows machines and 500 Macs with two people, but the Windows machines took so much of the time that he estimated he would need one full-time IT person for every 75 Windows boxes. The problem? Not that Windows was inherently harder to manage, but that with viruses, security holes, patches, and so forth, Windows machines required far more time to keep healthy. The MacOS and Linux machines required relatively little management attention.

All this has profound implications for Microsoft's business model. Unless Microsoft can both find a way to neutralize any competition, and provide a way to induce customers to get back on the regular upgrade train, its profit margins will inevitably shrink to a more normal level. This can happen either because of shrinking revenue (if Microsoft is forced to cut prices, or customers stop buying), or because of increased spending on R&D to develop new features and products.

In fact, the story could be even worse, depending on how Microsoft reacts to its maturing business model. At one level, we can look at Microsoft's business in the very macro scale and observe that (a) its operating margin is 2x to 4x the norm for large, profitable companies, (b) the forces which allow Microsoft to sustain this unusual profitability are breaking down, therefore (c) Microsoft's operating margin will have to drop by 50% to 75% over the next several years. That's painful enough if you're a Microsoft employee or shareholder.

Worse, though, is that Windows and Office together generate nearly all of Microsoft's profit, and a large number of unprofitable businesses (Xbox comes to mind) drag down operating margins. If this is true, then Microsoft will have a strong incentive to cut prices on Windows and Office even more dramatically to sustain revenue, while pulling the plug on unprofitable businesses. There is evidence that this is already happening , with rumors of effective price cuts of 95% (!) being offered to some foreign governments. It almost doesn't matter if the rumors are true: if large customers believe they can get Office for $25 a copy, that's what they'll demand, and if there's a viable free alternative, they'll get their price. Microsoft can probably afford an average price cut of 80% for Office and still have a positive operating margin on that product, even at current R&D and marketing levels. If the typical copy of Office goes for $250 today (after discounts), an 80% discount implies $50 a copy on average.

The truly worst-case scenario for Microsoft would be if competitors to its core products became not just viable alternatives, but actually started winning meaningful market share. That could force Microsoft into a situation where there's no way it can sustain Windows and Office as profitable products at current R&D and marketing levels. In order to save those business lines, Microsoft would have to dramatically streamline its R&D efforts, cut the products back to eliminate features which nobody uses (how many engineers do you you suppose it takes to keep Clippy running in a new version of Office?), or even stop major new development in recognition that new versions rarely induce people to upgrade anymore.

All this paints a dire picture, but there are still several things which Microsoft can do. When you're sitting on billions in cash, and adding over a billion to the horde every quarter, you've got a lot of options, even if the gravy train is likely to go dry over the next decade.

Strategy 1: Embrace (and extend) open source. If Bill Gates stopped by my house for grilled Ahi and asked my advice, this is what I would tell him: It is foolish to waste development effort on things which provide no competitive advantage. The core functionality of Windows and Office are now readily available for free from open source projects, and a couple of different companies have built Windows-compatible environments on top of Linux for what is (to Microsoft) petty cash. You can't fight that kind of a market force. The next versions of Windows and Office should be completely new products, built on top of Linux and OpenOffice.

That way, Microsoft will be focusing its development efforts strictly on features (such as a better UI) which differentiate its products from the alternatives, rather than repeating the efforts of the open source projects. In addition, any new features incorporated into the open source projects would free up Microsoft resources for stuff which actually provides an advantage over open source.

Some care would have to be taken to make sure that Microsoft's products themselves don't become open source under the GPL, but that can be managed by keeping the Microsoft features a distinct layer on top of the base open source project.

Strategy 2: Kill the alternatives with cash. It is possible (though not certain) that if Microsoft's products were free, then interest in alternatives would wane. This is basically the strategy of competing on price alone, taken to its only possible conclusion. This could leave the open source alternatives relatively unmaintained and unviable. Of course, it would mean giving up a lot of revenue until those projects die, but Microsoft can afford to do that for a long time.

This is basically the suicide-bomber strategy. This is the financial equivalent of saying, "I've got $50 billion in the bank, and if I go, I'm going to take everyone else with me!" I don't think this is a good strategy, though it is the inevitable result if Microsoft continues to compete on price. First, there's no guarantee it will work. Second, even if it did work, it would create a bloodbath in the industry. Third, it would destroy market confidence in Microsoft. Fourth, even after the alternatives were dead, it would be hard to start charging again.

Strategy 3: Change business models. There are a number of different business models which may be able to sustain growth where the model of selling lots and lots of software licenses can't. For example: give away the software, and sell support contracts (a la Red Hat). Or, integrate vertically and start selling Microsoft-branded PCs (a la Apple).

This is appealing in theory, because it attacks the fundamental problem head-on (the business model of getting customers to fork over for upgrades and new licenses is failing). In practice, it isn't so simple. Any change in business model would be excruciatingly painful for a company the size of Microsoft, and most of the alternatives would yield less revenue at lower margins in the near to intermediate term. In fact, it is only when the traditional business model has completely failed (for example, if Microsoft unwittingly found itself pursuing Strategy 2) that this becomes appealing.

History is littered with formerly-great companies which grew up on a major change in technology, and which died completely or became shadows of their former selves when they failed to adjust to market realities. Indeed, in many industries it was not the pioneers which survived in the long-term, but the companies which took over when the markets matured and the pioneers couldn't change.

Microsoft's current business is not sustainable in the long-term. The company is too profitable, and the forces which allowed it to remain so profitable are breaking down. Microsoft's challenge will be to shift the changes in the marketplace to its advantage, rather than trying to cling to the illusion of being a growth company in a mature market.