The View from Outside: Notes from an Infidel- by Steve Den Beste

Last month I tried to give you something to laugh about; this month I'll try to give you something to think about.

I grew up surrounded by rainforests in Oregon. My mind-bent has always been towards a system-view of the world (which is why I became a systems engineer, I suppose) and I became interested in ecology very early.

A climax rain forest is a beautiful sight; wonderful to walk in, very peaceful, and it smells like nothing else on earth. But from an ecological point of view, the site of a recent forest fire is a lot more interesting. A fire isn't necessarily a world-ending disaster. A lot of terrain is opened up, and for anything between 100 and 500 years there is a major botanical competition going on.

Other things can also open up such opportunities: on May 18, 1980 we got to witness a particularly spectacular "opportunity" when Mt. St. Helens "opened up" several hundred square miles of the Gifford Pinchot National Forest. But fires are a lot more common.

After a fire, there are opportunist plant species which move fast and seize the ground; mostly grasses and similar ground-cover. For a few years they may prosper, but eventually bushes move in, then small trees, then large trees, and eventually Douglas Firs finish the competition and take over.

But though grasses and bushes cannot last forever, there are always new forest fires creating new open areas. So there are always grasses and bushes somewhere.

There turns out to be a close relationship between ecology and economics. (This is hardly a new observation: one of the main sources of inspiration for Charles Darwin was Adam Smith's The Wealth of Nations.) So sometimes it helps to take an ecological view of a market.

Generally speaking, in the software industry a product category goes through four phases: birth, adolescence, maturity and senescence. (Think of these respectively as grasses, bushes, trees and forest fires.)

Birth takes place when someone has a nifty idea for an entirely new class of product. Of course, more people think they have a brilliant idea than actually do. (One of my favorite aphorisms is "Don't base vast projects on half-vast ideas.")

During a short window, the company originating the idea has an effective monopoly on it. Eventually other companies notice and create competing products. Once this happens, the product category enters adolescence.

The main characteristic of an adolescent market is that no-one really knows where they are going. New features are constantly tested in the market, with some succeeding and some failing.

But eventually there is enough experimenting that the desirable limits of the category have been probed, and users begin to remark "If only I had this feature of this product and that feature of that product in a single package, I'd be ecstatic."

The emergence of consensus signals the movement into maturity. As an adolescent market proceeds, products in it tend to diverge, but in a mature market they tend to converge. That's because an image crystallizes of what the ultimate product in the category will be, and the suppliers begin to move purposefully towards that goal, rather than trying to differentiate themselves from their competitors. This isn't plagiarism or anything sordid, it is merely the normal ecology of a maturing market. All successful competitors do it.

At the end of this process competing products will be comparable in capability.

It is not exactly true that stagnation sets in at this point, but the pace of change certainly slows. And eventually the product category becomes obsolete and the market size starts to shrink. That marks the transition to senescence. This often happens because a new product category has appeared (not a lot of demand for FORTRAN compilers these days), but can be caused simply by loss of interest on the part of customers (no-one markets new text-only adventure games anymore such as Zork).

From an ecological standpoint, each of these stages require a different set of talents. Birth requires a stroke of genius and a lot of guts (Edison's "1% inspiration and 99% perspiration").

In an adolescent market, engineering is the star. Products prosper through innovation, which comes from good engineering.

In a mature market, since all the products become comparable, customers become concerned more with things like availability, price, support, documentation -- in a word, marketing. Engineering continues to be important, but innovation is no longer key; engineering's job is to develop stable and trustworthy products so that the marketers can sell them well.

Once a market reaches senescence, additional massive investment in the product becomes pointless; the market cannot be expanded or even kept level, so excessive investment is simply wasted. The key to maximizing profits is to reduce investment in both engineering and marketing. This is a balancing act: too little and the product will die too soon; too much is simply wasted.

It is the fact that each stage requires a different set of talents which makes this so interesting from the point of view of an economic ecologist (a term I just invented).

Most people understand why a company usually doesn't survive the transition from birth to adolescence. On a feature-basis, they usually don't keep up. What is less well understood is the transition from adolescence to maturity, with the consequent de-emphasis on engineering innovation. (It's not that they stop trying to innovate; it's just that the lode is pretty much mined out.)

Engineers in particular tend to think that an adolescent market can continue forever, and that great engineering will continue to be the star of the show. (Sort of like the aging actress who won't accept the change from playing Ophelia to playing Gertrude.)

Usually the total number of vendors shrinks dramatically in the transition to maturity, sometimes to as little as one (if for no other reason because of the advantages of economy of scale). And because the companies who get shook out are usually the ones who continue to emphasize engineering, engineers tend to think that something sinister has happened. "It just isn't natural for engineering to cease to be the star. Innovation should continue to be central forever. There must be a plot."

The spreadsheet market is now mature. There is little important difference between Lotus 1-2-3 and Microsoft Excel. The difference in their market success lies in how they are sold, not in how they are engineered.

Understand that the statements I've made are all broad generalizations, and there will be individual exceptions. In an adolescent market there will be customers who are price-driven, and in a mature market there will be customers who are feature-driven. But the majority will go the other way. Likewise, in a mature market there will occasionally be significant innovation, but such advantages will be fleeting.

The PC operating system market has made the transition from adolescence to maturity. The rules have changed and the requirements for success are now different. Bush strategies will lose to tree strategies in the long run.

It is difficult to pinpoint precisely when the transition took place, but if I had to do so I would make it the release of Win 95. Don't misunderstand me: it is not that I think Win 95 is the ultimate operating system, but it represents a movement towards a unified vision of the ultimate operating system -- the first step of a long walk, if you please. It shows that a consensus has emerged, because Win 95 represents a dramatic break from Win 3.1 with its changed GUI -- which as has been pointed out is much more similar to the Mac and to OS/2 than was the case for Win 3.1.

OS/2 2.1 and Win 3.1 were extremely different. Win 95 and Warp are more similar than that, and Merlin and NT 4.0 are even more similar (though not by any stretch of the imagination identical). I do not mean to imply that there is no longer any technical difference between them, but there is less difference than between previous generations of products. There is every reason to believe that in future the technical differences will cease to be important entirely, except to spec-sheet fetishists. (You know the type: the ones whose volume controls go to 11 instead of 10.)

The mark of a mature market is convergence, and that is what we are seeing. And the star of the show is marketing, not engineering.

The goal of successful marketing is to position your product so that customers think it is the best one to solve their problems. Unsuccessful marketing consists of emphasizing the wrong things. If a customer looks at your advertising and says "So what?", you've blown it. (Note also that advertising is only one small piece of marketing.)

In the PC operating system market right now, lists of reasons why Merlin is technically superior to NT are becoming uninteresting. Technical comparison is a bush strategy in a what is rapidly becoming a tree market.

Tree strategies are preload deals, investment in effective advertising, proper product targeting, aggressive enlistment of support from ISVs and retail sellers, publicity, and good relations with the independent media. Does this sound like anyone you know? More important, does it sound different from someone you know?

That's why Win 95 is outselling Warp, and why I think that NT 4.0 will outsell Merlin unless IBM dramatically changes course.

(Steve will be back next month if he can find his way out of the forest he's gotten himself lost in.)


For those who are interested here are some samples and some thoughts on the mail I received from last month's article.
Steve Den Beste dropped out of college in 1975 when he was offered a full time job and never looked back. Since then, he has worked on products as varied as logic analyzers, factory automation, wide area networks, medical equipment and robotics. He has been programming microprocessors since the days of the Intel 8008 (talk about stone knives and bear skins) and antagonizing folks in c.o.o.a since 1996.

[Index]  [ Previous] - [Feedback] - [Next ]

[Our Sponsor: ChipChat Tech. Group - 32-bit OS/2 text paging software and Sound Cards.]


This page is maintained by Falcon Networking. We welcome your suggestions.

Copyright © 1996 - Falcon Networking