Don’t Dump Your Data: Supporting Data Projects a Great Way to Boost ROI

Data tends to be a poorly supported and under-championed function at most companies. On its face, data-related tasks are pretty dull and tactical. So, it makes sense that people rarely stake a portion of their career or budget on improving it. Although fixing data management (DM) may be dull, leaving data practices poorly supported means that IT departments are distracted by data and less able to deliver their projects on time, on budget and on scope.

Here is what a company needs to be able to do with data if it’s going to do things like eliminate costly redundant applications or combine the databases of its acquisitions:

Data migration, the process of transferring data among databases, platforms and applications. This is important when various disparate applications need to rely on a variety of distributed databases in order to provide information to end users.
Data integration, the process of combining data residing at different sources so that end users and applications have a single view of the data. This is important when a project involves the combining of databases so that duplicative applications can be discarded.
Quality management, which ensures that the data transferred among databases and applications can be properly used and is correct.

If you think this is dull or that it’s ok to relegate these data tasks to the IT-basement of boring tactical functions, think again. When projects get into trouble, it’s usually because of data. Maybe the data quality was so bad that the data scope of the project had to be reduced by a third when it finally came time to test and deploy. Or, maybe developers spent so much time manually creating data mappings that they cut corners when it came to meeting the original business requirements of the project. Or, maybe the team preserved the overall survival of the project by delaying its delivery by another three or four months.

How Expensive is Bad DM?

It turns out that all of these fixes are a lot more expensive than properly supporting data from the start:

A company will make $200,000 if a project that will eliminate $2 million in redundant annual software costs is accelerated just five weeks as a result of better DM.
A company will lose the entire amount that it invests in any project in which team members are overwhelmed by data and build an application with such poor functionality that end users don’t adopt it.
A company will lose $300,000 if a project that will save $3 million annually by accelerating accounts collection is reduced in scope by 10 percent.

Why DM Matters

It turns out that when developers, architects, project managers, and business analysts have the right tools for automating DM tasks, they become a lot more productive. More importantly, proper DM tools shifts people from low-level tactical tasks, such as hand coding of data mappings, to higher-value strategic tasks such as analyzing and improving the quality of data that a project relies on. Here are some of the benefits:

Improved developer productivity. The automation of data-related architecture and development tasks means that less time is spent on labor-intensive tasks such as hand coding of mappings and reworking of applications because of data quality problems.
Shortened project cycle times. Spending less time on hand coding for data-related tasks shortens project cycle times and prevents data problems from disrupting the critical paths of the projects they support.
Better project deliverables. When developers can automate the migration, integration and quality control of data and accomplish these tasks with a standardized set of tools, they are better able to meet the original business requirements of a project as well as maximize the amount and quality of data in a project.