I’ve been reading and listening to John Dvorak’s comments on the PC industry since 1983 or 84, when I was in business school thinking about what to do after graduation. I was interested in the potential of the PC industry, and I found Dvorak’s column the most useful for thinking about its future. Those were the days when he used a lot of bold type to emphasize his points.

What I like about Dvorak is that he tries to get behind what companies are saying/doing and comes up with an interesting perspective on what’s really going on. That leads to some projections that don’t pan out, but are interesting nevertheless. He retains this style in what he says on the Twit podcast. I met Dvorak briefly at a trade show 15 years ago. He was wandering the low-rent aisles and came across our booth. We had an antique MicroVAX there as part of a demo, which piqued his interest. He asked a thoughtful question or two about what we were up to and then moved on, belying his cranky persona.

A few weeks ago, the Twit pundits were talking about nascent Blu-ray player shipments. Someone commented that Sony may have run out of inventory because they built an initial quantity of boxes before committing to future runs, so they could gauge demand and set production levels appropriately. Dvorak responded that that was nonsense, nobody does runs any more, they do continuous production. He said that when companies introduce new products they’ve done market research and know precisely what demand will be and therefore how many to build.

Whoa! That stopped me in my tracks. I thought back to the very early days of Ipswitch, more than 15 years ago, when I met with the head of our accounting firm to talk about budgets. His principal comment: startups can project expenses pretty accurately, but when it comes to revenue, they tend to be way off. Now of course that has quite a bit to do with the fact that they are just starting out, and don’t have a basis for predicting how much or when customers will buy, but it’s been my observation that that has remained true for us as we introduce new products, and stays true even for very large companies. (Projecting sales of existing products is much more feasible.) No matter how much market research a company does, revenue projections for a new product are in the end an educated guess, which can be wildly off the mark. After all, with new products you’re trying to get consumers to change behavior or learn something new. Anyone remember Microsoft’s Bob product?

I also recall that large consumer goods companies like Procter and Gamble and Unilever (who are the real experts in market research) run all sorts of regional and local tests and still end up being blind-sided by a customer reaction to a new product because there was yet another way of characterizing the market that they had not anticipated.

As Ipswitch considers new products to develop, we use a formal process for market research, do analysis, and then define and build a product. Each time we do this we build upon what we’ve learned about the process before. After all of that, though, what we’ve done is reduce uncertainty, not eliminate it. It’s this balance of research, innovation, resource allocation, and operational management that has led to 15 consecutive years of growth for Ipswitch, and which is driving our strategy as we go forward.

But if you know someone with a crystal ball, John, please send them my way. I’d like to hire them.

Posted by Roger Greene
digg this add to del.icio.us add to My Web Furl this page

Trackback Pings

TrackBack URL for this entry:
http://156.21.1.19/mt-tb.cgi/1397