EMA's report, "20 Hot Markets for 2007" should be coming out of production just at about the time you read this. It was an opportunity for me, along with the rest of the firm, to explore less-than-obvious approaches to capturing less-than-obvious markets with an eye to the future rather than a linear projection of the past. As such, it involved pretty much the whole firm and has become a sounding board for our best minds to look at "markets" in everything from service management to storage and security, to VoIP and out-of-band, and business service management.

However, as the prime orchestrator of the report, I had a chance to formulate some of the ground rules: explicitly not to look for conventional markets in conventional terms but to find hot areas of growth that typically involved implosions and conversions of technologies, more often than not with a cross-domain (network and/or systems, and/or application) flavor. The result was, as I had hoped, not tidy markets, as they ranged from the very small to the very large. Nor were they necessarily designed for market sizing. Markets that are defined to be exclusive taxonomies, so that, for instance, a vendor or product can absolutely fit in only one or the other - often do not reflect the real underlying forces of how this industry grows.

For my money, the most interesting markets in IT management software are messy markets, where technologies explode in multiple and unpredictable ways, serving multiple IT user populations at once, without regard for the old, Mesozoic (read age of dinosaurs) taxonomies. The interesting markets don't care about networks or systems, or even applications; they care about infrastructure-enabled services. And the critical markets will explode so polymorphically that they'll touch - for instance - change and configuration, security and compliance, and service assurance and service provisioning, with the same rhythmic heartbeat.

Let's take one of the easier targets for ridicule or confusion - or perhaps a bit of both: "network performance management." One would think that, on the face of it, "network performance management" is a perfectly sensible term reflecting those solutions that primarily monitor the performance of the network. And that would mean the performance of network devices individually and collectively. And wouldn't that make sense? Yes, it would.

Denise Dubie is senior editor with Network World.