When I wrote the EAI book years ago, I never really focused on the business case for data integration, other than looking at the money being wasted in hand coding data flows between systems. Indeed, it was in the single digit billions then, worldwide. I’m sure it’s in the double digits now, despite the fact that great data integration technology exists these days.

So, if the benefit of data integration is so obvious, why do you need to create a business case, defining the ROI to justify the cost of the effort and the technology? The fact is that most major projects require them these days, considering that IT costs are much more controlled than they were a few years ago. Indeed, a CIO friend of mine has his staff create business cases for any project over $10,000, thus they are creating a lot of business cases there.

Creating a business case for data integration is more complex than you would think considering that you must take into account the existing “as is” problem domain, and understand the issues in the context of the business. Then, you need to determine what business advantages data integration brings, and how those advantages translate into dollars.

Here is the process:

· Determine the “as is” state of the problem domain.
· Determine the logical “to be.”
· Define the inefficiencies of the “as is,” and put a dollar figure next to them.
· Define the efficiency gains of the “to be,” and put dollar savings next to them.
· Create the ROI model.

Determine the “as is” state of the problem domain, means that we’re going to take a realistic look at how the existing systems are functioning without data integration, or in many instances, without effective data integration. What’s working? What is not?

Determine the logical “to be,” is really about what could be if data integration was in place. How will the use of data integration more effectively support the core business processes? What does that conceptual data integration architecture look like?

Define the inefficiencies of the “as is,” and put a dollar figure next to them. Now that we know the “as is” and a clear vision of the “to be,” let’s look at what inefficiencies are there, and how much they are costing the business, monthly, yearly, and over a 5 year horizon.

Define the efficiency gains of the “to be,” and put dollar savings next to them. Considering the last step, this is easy. Just put a dollar amount next to the efficiency gains from the “to be” data integration architecture. Make sure to consider the softer issues here as well, such as the ability to provide better customer service if the customer data is integrated with CRM system, meaning that those working with customers have a single view of customer data and it’s up-to-date. In most cases, that increases customer satisfaction.

Create the ROI model. This last step is to form an ROI model that is well researched, logical, and grounded in technology and business realities. It’s been my experience that the ROI from data integration is 500-1000 percent over a 5 year period of time, but your mileage may vary.

You never know until you run the numbers.