« A New Way To Think About Requirements | Main | Articles about IT on Archestra »

March 30, 2006

IT Investment versus Business Returns

Christopher Koch's blog for CIO.com featured the post Why Defend This Metric? (Thursday, January 05, 2006). The ongoing debate about how to quantify the business value of IT stays open, and at this point it seems the difficulty in answering the question is mainly because of the way the question is asked. Consequently, the debate deserves a turn in which the analysis might force a rephrasing of the question.

If a single metric could justify an investment, it would nonetheless actually have to account for several things, but the logic of how it does that is the most important aspect. Let's drill down.

First, the goal is to arrive at a fact that allows acceptance or rejection. This is about tolerance. The fact being sought is a piece of evidence that a significant difference has been demonstrated that is rationally attributable to a particular contributor -- in this case, "a business difference due to IT". The evidence may range from non-existent to very strong. But until there is agreement on the recognition of both the type of difference and its significance, there is no actual "executive" attention to what is acceptable.

Assuming that a type of difference and its significance have been defined and agreed upon as the axes of reference, a range of tolerance can be declared. But there is no point in attempting the declaration unless criteria have been explicitly established for identifying the boundaries of the tolerance.

Assuming that the range of tolerance is properly established, events that can represent the difference in question would have to be defined, and each type of event would have to be describable in terms of dynamics that normally correlate with the event. Those dynamics might be ones that "cause" an event or ones that "allow other causes to generate" the event. Since the difference between being a cause versus a prerequisite can be quite vast, the association proposed between a given event and a given dynamic must be rigorously distinguished -- in terms of the role of the actors (means) involved in the dynamic.

The reasons that explain why the actors have the roles that they exhibit must be exposed -- to allow recognition of where any behavioral or influential regularities originate, as well as irregularities.

Finally, opportunities for directing those actors -- which is the same as to say "for employing those means" -- must be identified and catalogued so that the exploitation of those opportunities can be examined both historically and in terms of operational policy and methods.

The ability to consistently coordinate these levels of description with each other, and to monitor their coordination, is clearly more a matter to be discussed as "competency" -- and the importance of that competency is in how decisions about what to do with information technology finally translate into significant business differences.

The proof of competency is measured in terms of real-time responses to business challenges. If those responses can be expressed as "scores" (just like a double-play in baseball can be scored in "outs" or hits can be scored in "bases"), then there can be a situational perspective brought to the utilization of the means, and the way that the scores are appreciated can be associated with the money spent on the involved means. When a win seems to require lots of base hits, and investment has been made in assuring hitting competency, the correspondence of the investment to the situation is easy to accept. But failure to hit well in those situations means little return on the investment. Likewise, failure to even be in those situations means little return on the investment.

In this sense, measuring performance against situations is obviously the correct way to measure IT investment. By comparison, measuring IT costs against total business income is apparently an arbitrary thing to do.

Posted by Malcolm Ryder at March 30, 2006 5:35 PM

Trackback Pings

TrackBack URL for this entry:
http://www.malcolmryder.com/cgi-bin/mt-tb.cgi/232

Comments

Post a comment