The IT Infrastructure Library (ITIL) is being touted in many quarters as the IT department’s salvation. The framework aligns IT processes with business goals, brings order to dysfunctional IT departments and satisfies an array of pesky compliance requirements. All true if — and it’s a big if — it is implemented properly.
Many ITIL projects do fail. More failures are a natural consequence of the recent increase in ITIL implementation projects in North America. About 25 per cent of Fortune 1000 companies are implementing some type of ITIL process, said Stephen Elliot, research manager at IDC.
Europe, by contrast, is an old hand. ITIL implementations started there about 10 years ago, said Keith Millar, vice-president of product management at Opalis Software Inc., a Mississauga, Ont.-based developer of data centre operations integration and automation software.
About half of Opalis’s customers are based in Europe, and the company has been involved in many projects to automate ITIL processes in recent years. Millar believes North American companies can learn valuable lessons from the “scar tissue” of Europe’s ITIL failures.
There are three main reasons for ITIL project failures, he said. Scope creep is one that IT people are all too familiar with, and that sabotages all manner of projects. “Some companies engage in chin-scratching exercises with the entire company about what would be ideal in IT service delivery. Expectations get set too high, and then delivery is way below that,” said Millar.
Another related reason is what Millar terms the “ITIL is the answer” syndrome, often seen in companies that have a “purist” approach. ITIL recommends companies look for examples of successful IT departments to emulate, which drives some companies to excesses. “So they look for an example of good configuration management, then they look at one for the service desk, and so on, trying to emulate success in each ITIL sub-component. Eventually, the executive with the checkbook will ask, ‘Why are we doing this again?’”
Companies should hive off and implement the ITIL components they can easily justify to help the business, and avoid exercises that can kill budgets, he said. But the number one reason for ITIL failures, according to Millar, is human nature. “What we hear back from our customers is that the main reason initiatives fail is the expectation that humans will follow rules. I think we’ve proven in the last couple million years that, as a species, we hate following rules.”
The end result of ITIL is ultimately a bunch of rules and procedures that must be followed by humans, he said. But people have trouble following time-consuming procedures that inconvenience them, even if they know they benefits the common good. A classic example is a printer jam late in the day. The official procedure may be to notify the help desk and complete a trouble ticket. But a rushed user may simply restart the printer and resubmit the print job, which solves his problem — but which also cancels all the print jobs other users had in the queue.
In like fashion, ITIL creates good processes for IT staff to follow, but many are repetitive. When companies shifted from mainframes to servers, what they really did was put a lot of “human glue” in between those servers to keep them running, said Millar. For example, a systems administrator investigating a server error will use the same techniques over and over.
Automating those steps once the new and improved ITIL-based processes are developed removes the unpredictable human element from workflows, Millar contended. “Opalis’s core concept is that this can’t be done programmatically….You need a GUI system that can be accessed by business experts.”
But IDC’s Elliot sounds a cautionary note: “There’s some obvious truth to the argument that there are improvements when humans are taken out of a process. But you also need a level of trust that the product can actually execute equally or better than humans.”
Elliot points out there’s a lot of ITIL hype nowadays. There are vendors who are looking to propagate their tool sets into companies by positioning them around ITIL. Some of these products are reactive and use proprietary scripts or capabilities. Companies should do some careful assessment before making investments, he said.
“There are products that take human labour out of processes, but there are some that actually perpetuate it. The point of ITIL is to reduce infrastructure complexity, not increase it.”
Elliot believes ITIL failures will likely occur more often in the future, due to the poor project management that plagues many IT projects. An unfortunate casualty will be loss of trust in IT’s ability to align with the business. “With business managers, one failure and your credibility goes back to square one.”