OS/2 e-Zine! - Chris' Rant

Chris' Rant

- by Chris Wenham

Summary: Have the various Internet "Market Research" campaigns actually been hurting OS/2's chances in the long run?

Death Of A Thousand Promises

To you and me it should be crystal clear right now that OS/2 is not surviving as a general-purpose operating system. That is, all the time it tries to be a Jack-of-all-trades, OS/2 is suffering from the effects of being spread too thin. For OS/2 to continue survival it must find a way to specialize. Now while IBM have their ideas for a niche, they are vague and constantly shifting. To me it seems that OS/2 has inherent talents that would make it perfect for a narrowly focused task. Under ideal circumstances the market would find the tasks to fit the talents easily -- sort of like tossing iron filings onto a table to see where the magnets have been hidden underneath. The problem is that the market hasn't been allowed to speak. Because some people are throwing paper confetti, and everybody is pretending to be a magnet. I'll explain.

In past months, several companies have put out "feelers" to test the potential of the OS/2 market. They do this by asking anyone who's interested in buying a proposed piece of OS/2 software to send them a request, which the company naively considers to be a potential order waiting to be filled. What they get, inevitably, are thousands of requests from seemingly eager OS/2 users who promise they're truly interested in buying the program. From that response the company then works out what the pricing structure and marketing focus should be.

Months and months go by, and then one day the company, having been charged up with such a positive response, releases the product that is everything it was promised to be. It sells a couple hundred copies in the first couple of months, after which the graph plunges to a miserable trickle of sales. What happened?

The obvious; thousands of the people who promised support for the product were being less than sincere.

"Well, Duh!" you say. But of course, it's unreasonable to expect everyone who promises to support a product to actually come through with the cash once the product ships. Maybe some of them lost their job or had unexpected expenses, maybe a competing product hit the market first, and maybe some just changed their minds. But for OS/2 the numbers are way out of proportion. Who wants to bet that Opera software will really see 10,000 orders as a result of Project Magic? In fact, does anyone really think they'll see more than three or four hundred at most?

We've been guilty of believing our failing support is producing the illusion of a lively OS/2 market and vendors have been guilty of thinking that real market research can be conducted over the Internet. Both parties are wrong. When Opera conducted Project Magic, they turned their market research into a game, one where the objective for OS/2 users was no longer to show a realistic measure of interest, but instead to beat the Mac and Linux users' scores. Now what will happen when Opera/2 is a colossal flop? We'll see vicious testimony of an "unprofitable OS/2 market", the type that scares away other vendors.

This is what I mean by confetti - software that just cannot manage in the OS/2 market anymore. When it's confused for the iron filings -- the software that can thrive -- people get the wrong idea when it won't stick to the magnets. Likewise, when everybody pretends they're a magnet -- be that a games player, a web surfer, a desktop publisher, a Java developer or whatever - just to make it seem as if there are a lot of magnets, then even the people throwing the iron filings cannot tell what is and is not working.

Thousands of requests, but only a handful of orders. Those profit margin estimates had been driven way off by a vaporous response, so what could have been a modestly profitable venture turned into a disaster instead. You see, sales don't have to be in the tens of thousands to make a product successful - just sales to the right people at the right price.

OS/2 users must resist the urge to sign up their support for any of these widely publicized "market research" dramas, because polling millions of self-selected web surfers is not the way you do basic market research. What's good for an opinion poll is not good for a profitable venture.

Send an e-mail instead that explains why you think the method is wrong. This way the vendor won't think his poll has worked, but is merely registering zero interest.

Register your interest in this column by joining in with our Hypernews Forum. If I get at least 10 replies I promise to write one more.

Chris Wenham is the Editor In Chief of OS/2 e-Zine! -- a promotion from Senior Editor which means he now takes all the blame.