Why IT Projects Fail

Computerworld - It's estimated that over 50% -- some say 80% -- of all large IT projects fail. The reasons for that startling failure rate aren't mysterious. Too many companies make the same mistakes too many times. I've identified five of the most common causes for failure. I've also come up with some questions your organization should answer before it starts a project, to boost the chance of success.
1. The project's real value isn't understood. When calculating ROI, be sure to have a clear understanding of the total cost of ownership. Without a concrete TCO, the ROI figures are meaningless.
Understanding the various ROI metrics, acknowledging that these are fuzzy metrics and clearly defining what the organization expects from a successful implementation will produce a clear view of the project's value.
2. You don't know your users. Who are the main users of the technology you're implementing? Are they business users or technical users, power users or casual users?
Business users are usually from marketing, finance and other nontechnical groups in the organization. They want to see summaries, dashboards, charts, graphs and other reports that enable them to spot trends.
Technical users are usually from IT, security and other operations groups. These are the ones who want all the nitty-gritty details. They want to drill down and diagnose problems. They perform the root-cause analysis when things go wrong.
Power users want sophisticated interfaces and flexibility. They generally use the system three or four times a week or even every day, and they can be anywhere in the organization, including marketing and finance.
Casual users are on the system once a week or less. They may not even log on at all and prefer to receive reports via e-mail.
3. Requirements aren't clearly understood. The only way to identify real requirements is to talk to potential users and get the answers to these questions: Are your requirements based on regulations, operational problems, performance, capacity optimization or trend analysis?
Do you have platform requirements -- Windows? Linux? Java? .Net? How about requirements to integrate with corporate infrastructure?
What's your requirement on performance? Scalability? Extensibility? Manageability? Usability? Security?
Before evaluating any technology, answer those questions in detail. Don't choose a product or create unreasonable requirements because of brand loyalty or FOE (friends of executives).
4. You've limited your options. Too often, companies decide that only a commercial product will meet their requirements, before all the options have been evaluated.
Before you jump to a commercial solution, you should evaluate your resources and skills within