Three ways to reprioritise your IT infrastructure investments
- 03 March, 2010 10:50
I've had the opportunity to visit several infrastructure and operations industry events already this year. Each targeted the same basic message: a return to innovation. But I have to wonder, are we really ready for innovation? The last 18 months have been marked by a consistent and constant effort to demonstrate return on investment (ROI). Why, as we now enter 2010, is there a sudden emphasis on innovation? The answer is what I refer to as "new shiny object syndrome."
As an infrastructure and operations professional, you're used to focusing on new, innovative uses of technology to support the business. However, the last 18 months have been anything but innovative. This period was characterised by project budget freezes and 20 per cent cuts to your operational budget. In fact, infrastructure spend suffered the deepest budget cuts in 2009, with a 16.3 per cent reduction from 2008 levels.
But all of that changes in 2010. Why? Because infrastructure will see the strongest IT budget growth. Our latest data shows that infrastructure spend will grow 9.2 percent in 2010, which is the largest overall growth and well above the 6.6 per cent average of other IT products and services. As a result, infrastructure spend is not only receiving the most emphasis in 2010, but shows the healthiest year-over-year delta at 25.5 percent.
Therein lies the dilemma. That's right, I said dilemma. It goes back to the shiny new object syndrome. As today's VPs of I&O see a return in their budgets they need to prioritise efforts and resist the temptation to go after shiny new infrastructure initiatives, like cloud computing.
The most important effort is to make sure the operations side of your house remains in order. Most infrastructure & operations departments are pretty good at focusing on operational efficiency by implementing frameworks like ITIL and deploying millions of dollars in IT service management tools from vendors like HP, IBM, and BMC. But that's not enough. Instead we recommend that you focus on three basic steps:
Step 1: Put the "O" back in "I&O" with IT process automation.
IT process automation is needed to take I&O to the next level of efficiency. Process automation isnt about reducing labor; it's about diverting resources so that staff that spends 75% of its time fighting fires can spend more time focused on business operations. We refer to this trend as IT Industrialisation.Why this is a big deal in 2010: Industrialisation is fueled by three trends culminate in 2010: 1) the maturation of IT operations frameworks like ITIL and Six Sigma; 2) the increased use of more sophisticated IT process automation tools; and 3) an appetite for organisational change.
What you need to do: It's often hard to change people and process, but in this post-recession era you'll find it's easier to take advantage of the crisis we're coming out of and realigning your teams' attitude, behavior, and culture. Stop rewarding your team as heroes that put out fires in IT and start motivating them to improve efficiency. The goal is to integrate IT operations with broader business operations. If you do nothing else in 2010, focus on IT automation tools to build a more efficient I&O foundation.
Step 2: Revitalise core datacenter gear, starting with the network.
After getting your IT operations house in order, you can move to optimising the datacenter. Use the early wins from IT automation -- often 40% in operation savings -- to fund your architectural refresh in the datacenter.
Why this is a big deal in 2010: Most firms update core datacenter gear every seven to 10 years. Chances are your last major upgrade was in 1999 or 2000 when times were good. Now we're 10 years down the road and the recession -- which temporarily delayed data center projects -- is now thawing. Move beyond consolidating and virtualising server infrastructure and demonstrate that similar optimisation efforts can be executed throughout the IT stack.
What you need to do: Rather than focusing on shiny new datacenter gear, work on collapsing tiers of the network by investing in new 10 GigE switch platforms; increasing the ratio of virtual machines to physic server ratios beyond 25:1; and finally building out new converged infrastructure that focuses on server, network, and storage in a single chassis. Look to IT heavyweights like IBM, HP, and Dell to deliver the most compelling solutions, but don't overlook new datacenter plays like Cisco with its EMC/VMware alliance. If nothing else, start your datacenter optimisation with your network. Your "throw bandwidth at it" upgrade strategy has outlived its usefulness and you need to revitalise an atrophied network fabric.
Step 3: Transform your desktop environment.
Finally, after improving operations and re-architecting the datacenter, 2010 will require you go after at least one shiny object: client virtualisation. But that's just the beginning; the end goal is to transform the endpoint.
Why this is a big deal in 2010: I use the "transformation" term lightly as it's often overused in IT. But 2010 is witnessing a confluence of events: the consumerisation of IT where employees are demanding that their netbooks, Macs, and smartphones are supported by IT; the rise of social computing as a legitimate business medium; the improvements in mobile broadband; and the shift to younger, tech-savvy workforce demographics. This is the year to rethink your desktop strategy and how you'll make your users, partners, and even customers more productive.
What you need to do: Transforming the desktop environment isn't as easy as it sounds. First, you must accurately segment your workforce. The two-size-fits-all days of providing either a desktop or laptop are long gone. Now you need to look at mobility and productivity needs to drive the architecture. Building a virtual desktop foundation is necessary, but you need to work closely with your security teams to revamp acceptable use policies. Also, chances are your remote worker policy is outdated too. To truly transform the endpoint you need to work with latest from Citrix, VMware, and Microsoft to support access from any device to any application.
Seeing a return to your I&O budget is a good thing. The challenge will be prioritising where you spend your time, energy, and money. Cloud computing -- and other infrastructure "shiny objects" are clearly the future -- but they often take three or more years to bear any fruit. Sure, you'll want to investigate these in 2010, but we feel the key to I&O is to 1) focus on industrialised models; 2) optimise the datacenter plumbing; and 3) enable new levels of user productivity. Use this three-step prioritisation to set your post-recession agenda.
Rob Whiteley is a VP and Principal Analyst at Forrester Research