« Needs versus Requirements: the battle of Why versus How | Main | Performance Improvement requires Change Management »

July 22, 2005

The ROI is dead; Long live the ROI

It makes perfect sense that CFOs would be the ones to consult about what is acceptable in measuring IT's economic impact.

Ultimately, the CFO determines the pace and level at which IT is funded.

This puts the CFO in a role that handles IT the way a real estate developer handles a commercial property. The expected value of the property is based on the utilization of the facilities and services that the property, appropriately prepared, will provide, protect and host.

That is, the impact of the "whole" property is taken as a new factor in the commerce of the area, while the commerce is the vehicle in which arrives the economic justification for approving the property. Ultimately, what is measured is the value of the commerce, with the property being simply a key enabler.

But that commerce is meanwhile constrained by zoning, competition, legal changes and weather. Properly acknowledged, these constraints are part of the considerations in the design of the property, but the property is not really expected to control or eliminate those constraints.

Let's look into what aspects of the CFO's view on IT are analogous to the above.

Some costs are purely about acquiring the property.
Acquiring the property is followed by preparing it.
Regulations and ambitions are negotiated into a development plan.
Construction and security commence.
Marketing is initiated in due course.
Occupancy is cultivated through rights and contracts.

Those steps cover the costs of creating the property and allocating/distributing it to users -- in other words, all of the asset deployment. Therefore, to package this group of costs even more tightly, let's call this group the resource delivery costs.

The upcoming financial issues are about the balance of maintenance and other upkeep (renovations, additions) versus income taken from the occupants. But from the developer's standpoint, that income need only "pay off" the delivery costs, and it might also pay back more than those costs.

Delivery provided a resource for the occupants. The occupants' situation was not to create the resource but to access it and utilize it. In the worst case scenario, the entire group of occupants fail to generate a net positive cashflow from their utilization. This case does not negate the fundamental potential for the property to return financial benefits greater than delivery costs. If all of the occupants were evicted, the property could be sold to a different management group, for the amount of its estate value, which is in effect its market value.

By analogy, IT has an estate value; it has inherent functionality that can be put to many different uses.

But as "developed" property, the IT infrastructure represents the managed configuration of the property, which underlies (encourages and supports) the benefits from its usage. Thus the managed property configuration has what I'll call the projected maximum resource value.

In real life, because occupants (users) and property managers are quite variously effective in their handling of the property, their changes can cause the equation for the final impact on the value of the commerce to deliver different answers.

Consequently, there is no directly causal relationship between the resource delivery cost and the commercial potential -- and obviously then neither between the asset and the final economic impact.

There is, however, a logical relationship. Logically, what should be understood and tracked is how the asset functions in the dynamics that result in the final economic impact. Initially entering those dynamics, is the asset an inhibitor, catalyst, accelerator or what? That is, what does the asset "do to" the outcome of the formula for final economic impacts?

The question is meaningless until the formula is available. Focusing on the dynamics must precede assessments of the assets. The dynamics are described at minimum in two dimensions.
- One: managing the purpose of resource utilization, and managing the actual utilization itself, is layered on top of managing the resource delivery and managing its underlying development. The logical linkages between these four layers of management are one dimension of the dynamics.
- Two: the other major dimension is the set of opportunities and constraints on each layer. Those factors determine what each layer can offer to its linkage with other layers.

The general challenge is to identify and exercise the controls that establish linkages and that select the opportunities and constraints actually being addressed in the linking.

Final economic impact is going to be the result of the throughput of these interlinked layers -- as applied to the prevailing conditions into which the final output arrives.

Accordingly, predicting economic impact involves assumptions about future states at each layer of management in the dynamics.

This means that the final representation of so-called "financial performance" (ROI) is a probability of a value.

Most of the attention to the economics happens when something is proposed that will ask internal workforce or work-partners of the company to:
- modify the infrastructure (engineering)
- modify the control of the infrastructure (maintenance),
- modify the ability to change the infrastructure or its control (management),
- modify users' leverage of the infrastructure (administration)

Keeping in mind that we see the infrastructure as a resource, the following diagram's framework of considerations represent coverage of opportunities to track the logic of potential economic throughput:

Financial and methodological commitments, which can support each major assumption about value throughput, should be called out and reality-checked for the opportunity and limitations that their current states present. These are, for better or worse, the controls on the throughput. By making them explicit, the overall dynamics of the investments become clarified as a set of requirements whose fulfillment can be measured. In this sense, achieving the desired throughput is a visibly active performance on the part of the company.

Posted by Malcolm Ryder at July 22, 2005 11:04 PM

Trackback Pings

TrackBack URL for this entry:
http://www.malcolmryder.com/cgi-bin/mt-tb.cgi/83

Comments

Post a comment