A while back, Jonathan Swift wrote "A Modest Proposal," an essay in which he suggested a somewhat unorthodox solution to the problem of famine in Ireland. No one took him seriously of course, which may be just as well. I'll admit that particular modest proposal may have been a bit unpalatable. Hopefully, what follows is not.Describing markets accurately is not a job for the faint of heart. Particularly when it comes to the storage business, terms such as "SMB" (small and midsize business) or "enterprise" often go unchallenged during attempts at market segmentation. Small businesses are usually designated "small" because the revenue they generate falls below some essentially arbitrary limit (say $50 million), or their employees fall below a certain number (100 perhaps, or 1,000).In fact, as I often try to point out to my clients that are vendors, these traditional methods of understanding markets really do stink. I have seen a few (very few) examples of vendors that try to understand potential customers in terms of how these companies use various technologies (I ultimately believe in something akin to this), but alas, they do this in a very inexact way and I think miss the boat in their analysis more often than not.For example, most vendors are not particularly clever when it comes to realizing that some very small companies use technology in a far more knowledgeable and sophisticated fashion than do many of their larger competitors. And most vendors are not too sharp when it comes to making a determination between what sales prospects want and what they will really get value from buying.Shame on those of my clients who do not listen to me. If use of technology is to be a yardstick, let's find something about technology that is more basic and more easily identifiable.It turns out that one of the shakiest dividers between markets is the one that exists between the "S" and the "M" in SMB. What is the distinction? In storage, it's an easy one, offered here for free (if you want more, my contact information can be found below).Right now, the SMI-S specification helps automate the discovery of arrays, switches and other devices on a storage-area network. SMI-S has been with us for a while now, and every serious player in enterprise storage builds toward this specification. In complex environments - large ones, or those smaller ones where on-demand computing and on-demand access to storage are taking over - nothing can be automated without the standards-based infrastructure that automated discovery requires.And yet, in some cases a standards-based approach is completely unnecessary. If your set up is a simple one - if for example you have only one array on your SAN - and if the SAN is in a rack located within a few feet of your desk, you will manage with near perfect efficiency if you just walk over to the device and see which lights are blinking and which are not.Clearly this is not an option for larger sites, but for smaller sites it should work every time.Thus, a good rule for defining one segment of the market - and one that functions quite well in terms of storage technology - is as follows: IF A SITE GETS NO ADVANTAGE FROM STANDARDS, TREAT IT AS A SMALL BUSINESS.If you are a vendor looking at SMB you find this offers a useful distinction between "S" and "M" -\u00a0and provides the granularity your approach requires.I think I'll consider making this and its corollary (small businesses get no advantage from standards) Karp's Third Rule of Computing.This method offers the advantages that no animals were used during testing, no children get eaten as a result of its implementation and, oh, yes, it does work.