Defining "IT Governance"

Got some feedback recently that the way I'm using "IT Governance" is "nonstandard." The point of view is that the term should be reserved specifically for the allocation of decision rights in IT (i.e., how the CEO, lines of business, and centralized IT should interact). This definition is clearly seen in Peter Weill and Jeanne Ross's recent book IT Governance: How Top Performers Manage IT Decision Rights for Superior Results.

Extending the IT Governance term to include such matters as IT service management, portfolio management, and the software development lifecycle was seen by this person as, oh, I don't know, imprecise and there was an implication of "how dare you not use the term exactly as Weill and Ross do." (Wasn't very positive feedback.)

Well, since my B.A. was in Political Science I guess I have at least a little credibility in such matters. Let's draw an analogy to the US Government. In the challenging point of view, the study of "U.S. Governance" would be restricted to only cover the power relationships between the exectuve, legislative, and judicial branches, and between the federal, state, and municipal levels. Discussion of how governmental entities establish and control programs to regulate behavior, provide services and collect information would be out of scope, as would discussion of how such activities interact with the political decision rights.

I think most if not all political scientists would regard this scope as impoverished. My view is that "IT Governance," like the study of human governance generally, may start with decision rights (authoritative allocation of resources, in strict poli sci terms). But to understand how those decision rights have meaning and operate effectively within a given context, one must start to cover (at least at a high level) governmental program management, regulatory structures, service provision, and information gathering - the "how" as well as the "what."

A simple example: the US Constitution calls for representation based on population. This is supported by the US Census, which makes understanding the US Census (and  maneuvering around redistricting) at some level important for understanding how political representation is actually working in the US. Similarly, in order to understand the effectiveness of IT Governance, one must understand how the IT organization (considered broadly) is structuring its major initiatives, control structures, and client relationships, and how effectively it is collecting and analyzing the information it needs to do its job (this last being a key topic in this forum).

So, "IT Governance" as used by erp4it (and corresponding writings) will continue to be a bit broader than the canonical Weill/Ross definition.

Thoughts?

-Charlie

Comments

[M. David Allen]

I have a few comments on definitions. There are a few guiding principles in this area, but the definition you're talking about has several issues mixed up in it.

First, the preference is always to use the canonical definition of a word. If you use a word to mean something other than the canonical definition, then it's incumbent on you to define what you mean up front so there is no confusion. (In this situation, it's not clear that Weill's definition is "the canonical definition" so it's hard to say if you're using it correctly)

Second, the term "IT Governance" in anybody's definition covers multiple areas. That would tend to indicate the better generic definition is the broader one, encompassing everything that could be considered "IT Governance". (That would be your definition) Sub-areas or sub-classifications such as Weill's, or something specific like "portfolio management" would be differentiated with additional words. (Like "Portfolio Management & Governance", or "Governance and Decision Rights")

Third, a definition is always tied to a purpose and a context. If you want to talk about governance from the perspective of a different use or understanding of it, it's of course legitimate to use a different definition. But it's still good practice to state at the outset what's meant by a term, particularly an aggregate term like "IT Governance" or "metadata".

The measuring stick is clarity of meaning. If your message is coming through loud and clear, then comments by pedantic definition people won't matter. If the meaning isn't clear, then definitions might be a problem.