Do you need to do ITIL? Wrong question
We can do without all this angst about "Should we do ITIL?" That is the last question we should ask ourselves.
[Updated Nov 2011: reorder questions to make them more customer-centric, "outside-in"]
The first question is: what does the business want us to fix? What is required for strategy and plans?
The second question is: what is actually broken? (Not meeting needs, not functional, sub-standard, too risky, needs improving). Just because it doesn't comply with ITIL doesn't mean it isn't working just fine: if IT ain't broke don't fix it
The third question is: is it worth fixing? Will we get a decent ROI? ITIL is irrelevant to this question. The ROI is measured by what happens before and what happens after, in business terms.
The fourth question is: is this the best possible use of the money, given limited funds? Where does this rate in the overall portfolio of proposed projects?
Only once you have an affirmative answer to all four of those questions should you then look at: how are we going to fix this? As a SMALL subset of that discussion, the question comes up: should we use ITIL as a tool, a reference framework, a benchmark?
It is only a small part because engaging the people who perform the processes, looking hard at the processes, and making changes are the key activities that bring the results. How does one distinguish ITIL process improvement from process improvement? I reckon you could use astrology as your framework for process review and improvement and you'd still show a significant result.
What framework you use for that transformation is of less importance than what experts you get in to help, what cultural change methodology you use, what executive support you get, how real the project is, and so on. All those decisions will help determine whether ITIl is useful or not in your case.