Playing the Numbers Game in Strategic Software Decisions

Numbers, numbers, numbers. Sometimes they are accurate. Sometimes they just fuel fear, uncertainty and doubt (FUD). FUD may be spread in different ways by different people, with different agendas. Consider the e-mail I received from a director of information technology today. It seems that his senior management wants to look at Microsoft Exchange for messaging. What propels this action? Two senior people in the organization spouting that Lotus Notes makes up only 5% of the total messaging market. Now I don't know about you, but I do not even remember the most biased industry analyst reports coming out with an absurd number like this. If this were true, it would mean that there are 2.36 Billion corporate messaging seats in play (118,000,00 Lotus Notes seats, as reported by IBM, would be 5% of this number). It appears these senior people either have an agenda or have been given really bad information from somebody.

Then there is the survey put out by Lucid8, a disaster recovery services firm. The survey purported to show convincing information that disaster recovery and downtime is a huge problem for users of Microsoft Exchange. Of course, they sell disaster recovery services for the Exchange marketplace. Ed Brill of IBM thought the survey numbers "seemed like a pretty good sample size and solid methodology", so he commented about the survey results on his blog, starting a spirited discussion, as these debates often do. But there is one problem. Lucid8 did not identify how the sample population was derived, and what survey methodology was used. In this case, it is possible that Lucid8 just published the results they wanted to sell their services, and Ed just bought in. Given the experience the IBM Lotus Software community had with a statistically flawed survey put out by an independent analyst firm, perhaps Ed should have not been so quick to jump on the potential FUD bandwagon here before getting more specific background information.