Tuesday, October 16, 2007

Could monopolies be healthly for the software industry?

Reading commentary about Oracle's BEA offer made me wonder if monopolies really are bad for software consumers. Logically, monopolies are detrimental in every industry because a single supplier is able to control prices that customers must pay. But there are some cases where a monopolistic structure seems to be not so bad or at least a natural result in certain industries.

In software, there are only two real factors in a purchasing decision: price and features. Price isn't just the amount that goes into the software company's pocket, but also the cost to implement and maintain a system. For large systems, the cost to simply train users might dwarf all other costs combined. As Microsoft has taught us, the biggest company tends to win out when price is the primary factor if only because training costs can be minimized. Nobody bothers to mention "Microsoft Windows" or "Microsoft Office" on a resume anymore, because every halfway qualified candidate has learned to use those programs already.

The other factor is features. Since the biggest companies have a huge advantage on the price side of the equation, upstart companies must compete on features. In my experience, it's fairly difficult to justify spending more on software on the basis of "nice-to-have" features. So in order to compete with bigger competitors, a small software company needs to create functionality that is so totally different and useful that its customers start to depend on it. For instance, a few years ago I purchased a copy of Quicken that downloads all of my transactions from my bank's website. Since I've grown to depend on this feature, Intuit has locked me into their software indefinitely.

The other lesson Microsoft has taught us is that big companies have an advantage when it comes to features as well, if the big companies catch the trend soon enough to copy the feature. For instance, Excel, Word, Windows, Money, Internet Explorer, and Outlook were introduced in order to outflank Lotus 123, WordPerfect, Macintosh, Quicken, and Netscape. There are dozens of smaller examples as well. Apple and Intuit survived only because they stayed under the radar long enough to lock-in a critical mass of customers before Microsoft moved in. Other competitors, such as Google, have thrived because Microsoft didn't understand their features until it was too late to emulate. Notice that these mistakes and oversights have occurred more often as Microsoft and the software industry have grown. It's just too hard for them to see everything that is going on.

From the customer's viewpoint, the Microsoft monopoly has been surprisingly benign. Sure, personal computers are probably too expensive because of the Windows and Office taxes, but cooperate America's software training costs are probably lower than they would be with more variety. It's hard to say if we are suffering from a lack of features, but until recently Microsoft has been the leader in distributing new types of software. Where they have failed, it seems like some other company has filled in the gap fairly quickly. In either case, innovation has thrived under the Microsoft monopoly to a greater extent than is possible to imagine under the IBM monopoly of the the 1970's.

Of course, once a monopoly develops, there is a new reason to buy software from a particular company: there is no other choice. And if everyone knows the monopolistic company will simply copy any new and revolutionary product, there is little reason for startups to startup. On the other hand, if the biggest companies are willing to buy up smaller players, like Microsoft in the 1990's and Oracle in the last three years, there is an incentive to fill functional gaps. From the market leader's perspective, purchasing successful competitors when they are small is both cheaper and more certain than developing their own copy. Customers also benefit, since the original products tend to be better than the imitations, at least for a while.

So the dynamics of the software industry may produce benevolent monopolies if:

  1. Big companies drive down the total cost of software ownership.
  2. Small companies have an incentive to compete on features and are not overly afraid of their ideas being copied by larger companies.
It's like a pond with two niches: small fish (that specialize in features) and big fish (that reduce overall price). Companies like BEA are in the uncomfortable middle: too big to be truly innovative and too small to be cost effective for customers. In this case, if the big fish swallows the medium-sized one, it might be best for the entire ecosystem.

No comments: