Friday, March 19, 2010

The Myth and Reality of ROI

Ultimately IT departments (and their clients) are concerned with a reduction in complexity, whether we are considering Identity Management, GRC, Information Security or any other initiative. This can measured in a number of ways:

  • Lowered TCO
  • Reduced Help Desk call counts
  • Increased usage of existing tools

No matter what metric is used, this is ultimately measured as Return on Investment, yes the dreaded ROI. Investment in software tools is always weighed against what this potential ROI can offer. Most firms specializing in Enterprise Software are more than happy to offer ROI calculator to prove the value of their offerings.

Customers and their Business Analysts consistently find that the ROI claims do not hold water. There are a couple of reasons for this, in my estimation:

  1. The ROI Calculator is defined too narrowly – Only a few parameters are highlighted in the calculator, and is therefore incomplete. For example, there’s no consideration for hardware costs, data costs, High Availability considerations, etc.
  2. The ROI Calculator is defined too broadly – The only way the numbers work is when they are applied too broadly. For instance, when a tool is needed for a workgroup or single location, but total enterprise numbers are the only ones that make it work.

Does this mean that the companies sponsoring the ROI are crooks and liars? Or those potential customers can’t do basic math? No, not at all. Just that everyone involved needs to be aware of how we plan to measure ROI. Vendors need to consider what the customer needs, while customers need to make sure that the offered metrics in the ROI calculator actually affect their organization.

When I first started outlining this entry is was more about technology and how the reduction in complexity would translate into required services whether they were based in the local data center or the cloud, however a discussion of the ROI benefits. The takeaway here is that once again business analysis is what drives the discussion.

From the technology side, the problem is fairly straight forward; do you use a simple monitoring solution that reacts to inform the client’s datacenter or a more modern intelligent system as advocated by Adrian Rodriguez and the team over at Likeminds? Along with a discussion of what the Enterprise’s requirements are, it will also depend on how these intelligent systems are evolving. It will be fun to watch, that’s for sure!

Post a Comment