Back in 2010, Canada’s then-auditor-general, the fearless Sheila Fraser, called out federal government IT for its aging, at-risk, inefficient infrastructure, and urged the feds to take a government-wide approach to dealing with these issues. The government responded in 2011 with its Shared Services strategy to rationalize IT systems and provide more bang for the buck by providing basic infrastructure services across departments.
There are lots of good reasons for pursuing that strategy. In 2013, according to the feds themselves, the government had more than 100 e-mail systems, most of them incompatible. There were more than 300 data centres, some of them doing the electronic equivalent of twiddling their thumbs while others struggled to keep pace with demand. Telecommunications networks weren’t co-ordinated. This wasn’t a result of poor planning. It was a result of trying to keep the country’s largest IT infrastructure running while architectures evolved and budgets flatlined. You can’t refresh an infrastructure the size of the Canada Revenue Agency’s every three years, and, by reputation, CRA runs a pretty tight ship,
There were also the examples of provincial governments taking the same path. The government of British Columbia started consolidating its IT efforts in 2002. By 2013, it had reduced the number of data centres to two from more than 100. Ontario says its consolidation initiative is saving $100 million yearly, about 25 per cent of its IT infrastructure spend. (You can read more about the benefits of a shared services strategy on the federal government’s web site.)
Federal IT soldiers along toward the shared services vision. As the B.C. example shows, it can be a long road. But the challenges that vision addresses, and the solutions it provides, mirror how a managed services approach to IT can benefit any enterprise infrastructure.
* Older infrastructure has to go, and not just because it’s old. Age isn’t the best reason to refresh your IT infrastructure. Anyone who has had an IBM AS/400 that hasn’t had to have maintenance for a dozen years can tell you that. Yes, operating systems might need upgrades to run the shiniest new applications, and processors start losing a step. But there’s a good chance an enterprise’s architecture, especially a mid-sized enterprise’s, just isn’t consistent across the board. Disparate offices made disparate infrastructure choices (especially with respect to telecom). Growth by acquisition makes the situation even more inconsistent. A managed services environment makes it easier to get branch offices singing from the same hymn book.
* A single infrastructure is cheaper than a dozen. Base-level services—e-mail, telecommunications, data services—can run from the same data centre, centralizing management demands.
* On a converged infrastructure, upgrades take place in lockstep. No office is left behind. Especially in a managed services environment, end-users might not even suffer any downtime. When processes change, those changes are built into the same window for every user, at the same time. There’s no need to convert the processes from Office A to match those from Offices B through Z because it’s running on the old application.
* Virtualization is critical to process efficiency. There aren’t many things more expensive than a printing press that isn’t running. But the press operator can change setup, change printing stock, change plates and—voila—the press is running a new job. Virtualization gives a managed services environment the same versatility. Rather than having some servers running at five per cent capacity, and some servers struggling to keep up, applications can be shifted fluidly to the hardware resource that’s best equipped to handle it.
The shared services model offers governments huge potential efficiencies. It also demonstrates how a managed services model offers similar benefits to any enterprise.