The cost of Canberra's poor IT procurement record

The blind test

The Department of Health and Ageing, for those that need the background, signed a $109 million deal with incumbent provider IBM Global Services for a further four years in late December 2010 to extend on a ten-year partnership. This was an extension of a $126 million extension signed in mid-2008 – that also avoided the open market process.

It felt fitting – given that the news was announced close to Christmas – that iTnews would spend part of January looking at this deal with a little more scrutiny.

So we sought a list of technologies that make up DOHA’s IT environment and presented them as a list of technologies in use at a large, unnamed organisation to some of IBM’s largest competitors (CSC, HP Enterprise Services and Fujitsu).

The aim of the exercise was to determine whether there were any technologies listed that they did not have the skills to manage - whether DOHA's IT environment was truly so unique that it should excuse itself from going to tender.

The competitors were presented with the following list:

Desktop:

• IBM ThinkCentre PC’s
• Windows XP (desktop operating system)
• MS Office 2003 (office productivity)
• Lotus Notes 7.0 (collaboration tools)
• Symantec Antivirus
• Adobe Acrobat Reader
• WinZip (file compression)

Data centre:

• IBM mainframe
• IBM blade servers
• Cisco Systems switches
• VMware (virtualisation software)
• SAP (financials)
• Some hosted desktop applications running Access and Excel

Keep in mind that nobody from HP, Fujitsu or CSC knew I was referring to DOHA – it was a blind test. Here were the responses from the CTOs at these service providers.

“The environment you describe below does not present any unique challenges to CSC. All of the elements you name are managed by us for different clients. Obviously, we'd need to know a lot more, particularly about any specific application requirements, but at face value this seems to be a portfolio of systems that most outsourcers could handle.”

“I can confirm that HP Enterprise Services offer solutions in all the areas on your list.”

“Fujitsu handles all of the technologies below (and more) for literally thousands of customers via various forms of service including systems integration, applications development and management, facilities management (onsite and offsite), through to total outsourcing, offshoring and cloud. A large proportion of our business comes from government and we are accredited at the highest levels of security for even the most sensitive Government applications. These capabilities all exist within our ANZ operations.”

I asked Ovum analyst Jens Butler to look at this list, to ensure some independence in responses.

"There is nothing unique about it, from a technology view. However, you need to consider the complexity of the environment, locations, type of services and how they're delivered to the business, number of versions of the applications running, the functions being supported, the type of business critical processes, type of skills employed to manage the environment etc."

I sent the same list to DOHA, asking whether there was any other infrastructure or applications requiring management that I didn’t include in the list.

The response:

“The Department’s requirement for outsourced ICT services is not based on equipment and applications. Detailed cost/benefit analysis, including benchmarking undertaken by the Department’s IT consultant, advised that direct sourcing provided optimum value for money for taxpayers.  The additional costs that would be incurred in going to market and the costs of potential transition to new service providers, while implementing the Government’s health reform agenda, were not justifiable.”

Butler (reminding you that his comment was based on a blind test) agreed that going to tender requires an internal governance team being in place or (in its absence) the hiring of third parties to oversee the process.

And to that I’d like to point out that DOHA had already spent $1.5 million on consultants to evaluate sourcing options with a view to going to tender (so much for wanting to avoid the costs of going to market). The relatively generic technologies involved would - at a high level - not necessitate a particularly expensive transition.

With this in mind I’d like to return to a question posed earlier.

How much of the $42 billion spent on contracts by agencies, $3 billion of it on IT and telecommunications, could the taxpayer expect to save if a more rigorous procurement process was enforced?