I’ve been on a terminology binge recently, defining terms such as machine-generated data, analytic platform, internet request processing, and transparent sharding. So perhaps this is a good time to introduce
Monash’s Third Law of Commercial Semantics
No market categorization is ever precise.
The reasons this is true may be flippantly summarized as:
- Bad jargon drives out good. That, of course, is Monash’s First Law of Commercial Semantics.
- Hard cases make bad jargon. I borrowed that one from the law dictum “hard cases make bad law.” Replace “hard cases” with “edge cases” and “make” with “lead to,” and you should see the point.
- Nothing concise is ever precise. This principle applies far beyond the marketing domain — for example, it’s why reasonable judges can reasonably disagree.
A more sedate set of reasons goes something like this.
A market category usually will have both core members and non-core. A faithful description of the core will fail to cover some outliers. But try to include the whole thing and you’re apt to be unhelpfully vague. Whichever choice you make, the categorization is imperfect. For example, my recent proposal of “Internet Request Processing” is awkward in those relatively few cases where wide-area networks are not accurately described by the term “internet.”
Market categories often merge due to deliberate product integration. This is particularly true in the software industry. For example, through the early 1980s, accounts payable, purchasing, and inventory applications were three separate things — but developers then noticed that all three should really be driven by the same vendor file or table, and it became rare to see them marketed separately.
Even absent deliberate integration, products often pick up desirable features from other categories. Examples may be found anywhere a “low-end” product adds a few “high-end” features. Others may be found in the way DBMS both old (Oracle) and new (SAP HANA) are extended in the hope they will be all things to all people.
The issues aren’t just on the vendor side; users’ technical strategy decisions may not map well to vendors’ product categories. In part, this is another version of the product integration point; users like to reduce the number of vendors they use, and hence may place exaggerated weight upon the alleged integration and completion of various product lines. Beyond that, as my consulting experiences consistently show, organizations interweave use cases in really interesting ways. Thus, user and use-case categorizations are no more accurate than product-oriented ones.
And last but not least, marketing claims are commonly exaggerated. Categorizations based on reality are somewhat different than those based on what vendors — or sales-driven analyst firms — claim.