CIOInsight has revealed the bitter truth (via InfoQ) about Return On Investment (ROI) for IT projects.
IT and business executives overwhelmingly agree: Their companies receive business value from their IT investments. But how much value, and what kind of value, is as clear as mud. When will there be progress?
One of the major reasons for this sorry state of ROI measurement in IT is that it is still at an incubancy stage. There is hardly any progress and even if there is any, it is localized. The most basic thing is to first identify the ROI points, only then can we measure them. Most of the times the measurement fails either because the right ROI points are not identified or they are not identified at all.
The important thing to consider in calculating the ROI is that it completely depends on the business that wants to implement IT rather than on the IT project itself.
ROI does more than just a couple of things. It
- provides a mechanism to perform cost benefit analysis to choose the best solution
- can provide a base for defining budget for the IT project
- defines a set of parameters that can guide the IT project. ROI is valid within certain parameters, e.g., timeline. This can be used to define the deadlines for the IT project
- helps in performing cost-benefit analysis and feasibility study for the venture.
There are two critical aspects of ROI - that IT is considered as an investment and secondly it justifies the effort and the temporary disruption. The complexity and accuracy of ROI depends on certain factors.
Is the need already established?
ROI is easiest to handle when it is considered to satisfy an already established need. In such a case ROI is directly proportional to the cost of not satisfying the need. A very simple but common example is having a web presence. Today a corporate website is a must, without it the business establishment is considered incomplete. The ROI of developing and maintaining the website can be identified by considering the cost of not having the web presence.
In some cases, IT is a core part of the business, e.g., in cases like ebay.com. Without IT probably that business would not exist.
However, there are some cases where need for IT is temporary, e.g., moving data or digitizing certain archived data and such. In such cases, ROI is usually an output of comparisons of different ways of performing that activity. For example, if text-mining is required to move to a database application it can done either using automated ways or by employing a certain man power. In such cases, other factors like labor cost and the timeline can matter.
What is ROI being used for?
ROI can be used for multiple purposes. One can be to identify if the IT is indeed a solution or not, and is there a better solution. Or it can be used for defining budgets for the IT project.
However, the common thing in all these is that the ROI should be able to point to the bottom line justification for the business. Even if does not give the numbers, it should atleast lead the way for the IT project. All the technical decisions in the IT should be able to be approved from perspective of the ROI. If they are not based on the ROI then they are not necessarily working for benefit of the business.
Are we targeting multiple problems?
It is not uncommon to use a single IT exercise to target multiple problems in multiple domains. In such a case, ROI should also include interaction and overlapping between these problems and domains. As they say, the whole is more than summation of the parts.
A critical problem in determining ROI in such cases is to get all the domain experts on the same ground at the same time.
Theory Of Constraints for identifying ROI
I have brushed this topic earlier. Theory Of Constraints (TOC) says breaking of an existing constraint can provide the maximum benefit. If IT is breaking a constraint, it is the easiest way of identifying the ROI. In such cases, the ROI gets more accurate.
From the other side, software builders can target at breaking certain existing constraints of businesses, thereby identifying and providing a direct ROI.
Measuring ROI
The biggest problem of measuring ROI is that it cannot be absolute, at best it is a judgement. It depends on information that is not readily available and is gathered out of surveys or predicted. Any of these can change, sometimes making the ROI invalid. In most of the cases, considering all the factors that affect the business should be considered in calculating the ROI.
Another problem is that there are multiple scales for measuring ROI, like customer satisfaction, convenience, accuracy, monetary, performance and probably many more. It is usually very expensive to normalize the different scales to get an overall ROI judgement.
However, at the end of the day it should translate into bottom line benefit for the business, that is, it should help the business achieve its goal in expected time at expected cost.
