(ROI) Return on Investment for Business Intelligence

How to justify a business intelligence system has occupied people ever since these were called decision support systems.  Supposedly the numbers are “soft”.  Reductions in IT costs virtually never cover the cost of replacing spreadsheets and Access databases with a proper BI system.  Yet, organizations spend over a billion dollars a year on BI systems, and that’s just in the USA.  In today’s environment, CFO’s will not let capital investments pass without some kind of a business case.  So what do these companies do?

Let’s briefly dissect the formula for return on investment:

ROI =     Expected value (NPV (cash flows from revenue increases or cost savings))                                          
…………Expected value (NPV(cash outlays from spending on hardware, software, labor, and services))

This formula tells us what we have to look for and how we should maximize the value.  We need to estimate:

• Cash flows and their timings
• Likelihood of these benefits occurring
• Cash outlays for the BI system and their timings, including
• Hardware and software costs
• Implementation costs
• Training costs
• Administration, enhancement and upgrade costs
• The likelihood of the cash outlays occurring

The formula also tells us that benefits and costs that occur in the near future are more valuable than those in the distant future because they are discounted less AND because the probability of them occurring is higher.

The classic definition of ROI looks at cash flows.  Non cash accounting items do not count.

Someone who wants to be seen as delivering valuable BI projects, therefore will look for a project that delivers a set of benefits that drive value over a short period of time.  If the scope is too small, the benefits will be minimal.  If the scope is too large, the risk and time to deliver will be unacceptable.

I understand this is not a new finding.  The formula just lays it out in black and white.

In future posts, I will discuss how to estimate these numbers, even when people feel they cannot be quantified.