The Texas Ten-Step

When a state tries to merge data centers from all its agencies, it's in for a bumpy ride -- and a very long one.
by | January 1, 2009

Texas is in an uncomfortable place, technology-wise. You may have been there, too. You take a huge and expensive leap on a big, complicated IT outsourcing project, and things go haywire. The local press gets hold of the news, and your state's IT laundry is aired for everyone to shake their heads at and tsk-tsk over.

Around the country, chief information officers are empathizing with Texas' pain. But they also are relieved that it's not them -- this time. And they're keen to glean lessons from Texas' difficulties.

A little background: Under a state law that took effect in September 2005, Texas agencies were ordered to hand over responsibility for IT infrastructure to the Department of Information Resources. The idea was the state could save money and become more efficient by consolidating 31 data-center operations into two and outsourcing them to IBM. From the get-go, the buzz in IT circles was all about how difficult it was going to be to get agencies to work together. Many believed that resistant agencies would try to get waivers, which could undermine the process.

Well, the $863 million, seven-year deal with IBM has indeed hit rough waters. Agencies have reported network breakdowns and failed server backups. Thousands of documents created by the attorney general's office were destroyed in July when a server crashed. Other agencies have reported lost data and multiple backup failures. The state has penalized IBM for its failure to complete backups on time.

State Representative Carl Isett, who authored the 2005 outsourcing law, has said no one was naove enough to believe there wouldn't be "hiccups." But those hiccups have become major complaints that have ricocheted within and around state government agencies about IBM's responsibility for Texas' outsourcing predicament.

At the same time, some say that whatever problems existed were compounded by hostility from state employees, and that despite everything, Texas is better off now than it was before. Technology processes are faster since the outsourcing began, and IBM is pulling and dragging Texas agencies into the modern age. For example, the state was using old machines that were on the verge of outliving their usefulness.

IT experts say it's too early to know how Texas's outsourcing will fare, but they offer general lessons and thoughts about outsourcing.

States undertaking a large outsourcing need to address, going in, how they will deal with staffing issues and knowledge transfer from employees. They must have a plan that takes into account which staff knows what information and whether those people will be part of the team working on the project.

It's easy to find villains in IT misfortunes, but diagnosing misdeeds may not be that simple. It's tough under any circumstances to transform government IT and run government at the same time -- especially if there aren't any extra dollars to spend. And a whole lot of agencies are expected to change over at once.

Outsourcing is difficult, period. Dozens of agencies that normally keep their IT cards close to the vest are asked to reverse direction and show everyone their hands. And then, everyone in those agencies is expected to be good sports about completely changing how they do business.

Public IT wreckage isn't good for anyone. It triggers a legislative cringe reflex that can stall other projects and make legislators balk at giving huge dollar amounts to technology for a while.

If a state hands over unmanageable legacy problems during an outsourcing, why does it think it will be easy for anyone else to fix them? Government does get someone else to point fingers at, but it hasn't necessarily mitigated the risk of revamping systems.

Many states are struggling with whether and how to consolidate. Gartner, a research company, does not offer encouraging news on this front. In a January 2008 research paper, it predicts, based on past evidence, that through the end of next year, fewer than 20 percent of governments will receive the intended benefits of a major consolidation or shared-services initiative.

Among Gartner's reasons for the high failure rate: ineffective or nonexistent governance, overly aggressive or unrealistic cost-reduction goals, insufficient investment and poor project management -- things even the most talented and charismatic CIOs can't overcome.

As to whether the Texas project is a success or a failure, few are willing to choose one or the other. Some come down on a middle way: It's neither. It's simply a long process.

Ellen Perlman
Ellen Perlman  |  Former columnist
mailbox@governing.com  | 

Join the Discussion

After you comment, click Post. You can enter an anonymous Display Name or connect to a social profile.

More from Tech Talk