Texas warns IBM on failed data center consolidation

In a classic case of the "blind leading the blind," Texas accused IBM of failing to deliver on an $864 million IT initiative.
Written by Michael Krigsman, Contributor

The Texas Department of Information Resources (DIR) sent IBM a "Notice to Cure," accusing the large system integrator of failing to perform its obligations on a data center consolidation contract worth $863 million.

According to an internal report prepared by the department, this is a case of the "blind leading the blind," with both parties at fault.

The press release issued by DIR includes sabre-rattling language from the DIR, as the department tries to force IBM into compliance with its requests:

The accumulated effect of under-investment by IBM, poor performance, and continual disregard for the protective obligations of the MSA, has resulted in harm to state agencies, exposure to unnecessary risks, and failure to achieve the objectives set and agreed by IBM.

The Notice to Cure is DIR’s response to repeated performance failures over an extended period of time, and follows months of unsuccessful attempts to find a collaborative solution with IBM.

The Notice to Cure gives IBM 30 days to remedy identified deficiencies. DIR expects this action to result in immediate and substantial performance improvements.

The Notice to Cure (I embedded the entire document at the bottom of this post) offers a harsh indictment of IBM's performance to date. It blames IBM for failing to meet contractual obligations in numerous different areas.

Here are the top six allegations:

Transformation. IBM has failed to complete Transformation Services in accordance with the terms of the MSA, including Exhibit 20. Specifically, IBM failed to complete Transformation of the twenty-seven (27) agencies into the consolidated data centers within the applicable timeframes. To date, only five (5) agencies are completely transformed.

Backup and Recovery. Despite the essential nature of backup and recovery services. IBM frequently fails to perform all required backups on systems under IBM management. including systems implemented in the consolidated data centers. IBM also fails to provide verification for all managed systems whether backups are in fact performed as scheduled and completed successfully.

Disaster Recovery. Significant deficiencies noted in the latest rejection included, among other things: runbooks and technical recovery guide sections are populated with boilerplate language not specific to the agency. the server or the application; applications installed and business functions supported on a server are missing from runbooks and recovery guides; recovery and dependency infonnation related to applications is typically not documented;. and recovery details associated with backup and recovery are not sufficiently documented to enable step-by-step recovery.

Staffing. IBM has failed to provide sufficient and suitably qualified personnel to perform the services, causing severe procurement, project and service request backlogs that have materially and adversely affected the ability of the agencies to serve their constituents and fulfill their missions. IBM has continued to remove staff from this account despite the growth of the backlogs.

Security. IBM has breached these contractual obligations by repeatedly failing to adequately track and monitor its activities associated with off-hoarding IBM personnel from this account. IBM has also failed to fully implement required Antivirus Software and to timely and consistently apply software patches for all managed systems....

Service Level Failures. IBM's chronic failure to attain service levels, disregard for required corrective actions and failure to implement the tools and processes necessary to measure all Service Levels places IBM in material breach of its obligation to provide the Services in accordance with the Service Levels.

According to Government Technology, IBM denies the allegations:

"IBM has fulfilled its obligations under the contract and today's action by DIR was unnecessary and unjustified," said company spokesman Jeff Tieszen. "IBM has worked in cooperation and good faith with DIR to provide benefits and improvements to all citizens of Texas. IBM very much regrets the state's action and will aggressively protect its interests going forward."

Strategic Analysis

This failure is another example of poor alignment between customer expectations and service provider delivery. Despite extensive project contracts and documentation, much of which is available publicly, the parties cannot agree on what is actually required under the contract.

Lack of agreement is a common source of problems on large or complex projects. Other recent examples include the EDS v. BSkyB lawsuit in the UK and Marin County's lawsuit against Deloitte Consulting. In both those cases, the enterprise customer claimed fraud against the service provider.

Related: EDS to pay $460 million over CRM failure

Related: Marin County sues Deloitte: Alleges fraud on SAP project

My take. I suspect the key to this case lies buried in an obscure review document commissioned by DIRS:

It is unlikely that either DIR or IBM could have anticipated this misalignment during the contracting process. DIR had little to no prior exposure to either large-scale managed services or complex multi-agency infrastructure projects. And IBM had limited experience in conducting such arrangements in the context of a state government environment such as the one operated in Texas.

In other words, according to the Department's own report, neither party had sufficient experience to implement a program of this type and scale successfully. Yes, it's a blind-leading-the-blind scenario.

Projects of this size are usually doomed to cost and schedule overruns. Nonetheless, the seduction of theoretical benefits causes otherwise rational people to go down the path toward failure.

What human tendencies drive us into situations of inevitable failure. Please add your voice in the comments!

Full text of Texas-IBM Notice to Cure

Photo from iStockphoto

Editorial standards