BeyeBLOGS | BeyeBLOGS Home | Get Your Own Blog

Main | July 2007 »

May 16, 2007

Data Management

I took a mini tour of the East coast last week for a conference and visits with clients. My multi-city trip began in Florida and ended in Washington D.C. I went to Florida for a DAMA meeting, then Baltimore for a meeting with a mutual fund company, then a mortgage lender (not a subprime lender!) and finally ended up at the IT Compliance Institute conference just outside of D.C.

While it was a bit of an eclectic trip, I couldn’t help but notice that there were two major topics that came up at every leg of my trip. They included:

1) Data Management and Organizational Design
Basically, how companies are structuring themselves to better manager their data empires.
2) Project Justification
How IT justifies to management the needed investment for IT infrastructure (software and hardware) to better manage data governance and compliance initiatives.

Today, I want to address organizational design issues. My next submission will explore the process for project justification.

Companies are struggling not so much with decisions about, “should I buy this ETL tool, or that EAI tool or some other data analysis product,” but more about, “if I did buy productivity software, how should I best organize myself to take advantage of it.” Actually, the thought process is in the opposite order. The first question they are asking is, “What is the best way to organize the data management part of my IT organization so I can be more efficient?”

An emerging trend that I recently seen in larger organizations is the centralization of data management groups to create greater leverage in training and use of technology. The concept started a few years ago with the introduction of Integration Competency Centers (ICCs) that were mostly focused around ETL tools. That idea is expanding beyond just ETL and seems to include all manner of data oriented projects. The companies that are successfully implementing and completing data intensive projects are the ones that are centralizing data management expertise and then leveraging that expertise across the organization.

There are several reasons why centralization is critical and they include:

1) Skill development. The tools available for data management are not simple. Databases, ETL Tools, or even tools like which my company creates—enterprise software that discovers business rules between structured data sources—they are not the kind of products you just pick up and start using 10 minutes later. So companies and employees have to invest the time to learn how to use these tools effectively. The problem is that if one group uses a tool one week, then the next group uses it for a week, but they both don’t come back to use it again for a month, no one in the company develops sufficient skills to use these new tools. So while the tools are designed to provide tremendous productivity gains, that can only be accomplished if you have proficient users. That means it is better to have a critical mass of users centrally trained as experts who are using a given tool all the time.
2) Project Leverage. By centralizing data management skills into a single group, it becomes much easier and much more likely that work done from one project will be leveraged to another. That is just because the centralized group is more likely to know that a similar project was completed elsewhere in the company. While you could argue that this could be handled by better communication, one way to improve communications is to simply reduce the distance and organizational divide between the people communicating. Hence, centralizing data management skills facilitates better communication around data and hence better leverage.
3) Buying Power. Often, one business unit can’t afford to purchase some hardware or software because they aren’t going to use the product in question enough to justify the purchase. By centralizing data management, you can not only leverage knowledge across projects, but you can leverage the software solutions available across more projects, which can significantly improve the return on investment for the solution in question.

Centralizing data management isn’t the only organizational paradigm out there and there are certainly real tradeoffs to be made when centralizing skills versus having them decentralized. However, it does seem that the companies that are completing data intensive projects on time and within budget are the ones that aren’t just throwing dollars at their problems—companies are first organizing themselves and establishing clear accountability for success around data management.

Next time: Project justification. How companies are improving ROI and justifying projects.

Share: del.icio.us Digg Furl ma.gnolia Netscape Newsvine reddit StumbleUpon Yahoo MyWeb  

Posted by Alex Gorelik & Todd Goldman at 1:45 PM | Comments (0)