Why Do Large IT Projects Fail?
Each year, the largest companies in the world cancel hundreds, or even thousands, of information technology (IT) projects before they are completed. These projects might be new application software (custom or off-the-shelf), new network architectures, or even new desktop operating systems, but they share one characteristic: they were discontinued by management before completion. The company never reaps the anticipated benefits and wastes thousands, and sometimes millions, of dollars.
In this white paper, we will examine the primary reasons that large IT projects fail, and we will outline some of the steps that you can take to avoid these pitfalls and give your next significant IT project its best chances for success.
Poor Definition of Project Scope
Incredible as it may seem, many IT projects are commenced before the scope of the project has been decided. Fundamental design and architectural decisions are made, hardware is acquired, and programming commences. After work is well underway and substantial resources have been committed or consumed, the scope of the project is finalized, and the feature set is fully defined. At that point, some of the earlier decisions must be reevaluated. Is the hardware adequate to support the anticipated number of users? Is the central database robust enough to handle the expected load? Are the end users running an appropriate operating system on their desktop computers? If the answer to any of these questions is no, hardware will have to be replaced or upgraded, software components may have to be replaced, and large amounts of completed work may have to be revised or even discarded.
Solution: Create a detailed project specification, and insist that the managers of all the groups or departments that will use the project review the plan and sign off on it before you begin work.
Lack of Input from End Users
Even if the project meets its specifications and is delivered on time and within budget, if the end users are not willing to use it, the project will fail. Users are often unwilling to use a new system if they do not understand it, it does not mesh will with their other systems (both computer-based and manual), or it fails to meet their needs. Systems are often developed with inadequate (or no) input from the end users of the system, so unless the designers are very much in tune with the needs of the end users, the system may have to struggle for acceptance.
Solution: Involve a good mix of the intended end users (for example, users with basic computer skills as well as users with advanced computer skills) in every facet of the plan, including developing the specifications, reviewing prototypes, and testing the software before rollout.
Lack of Upper Management Commitment
IT projects can be very expensive and take months (or sometimes years) to complete. In these days of global competition, tightened budgets, and downsizing, large IT projects are readily terminated, especially if they are perceived as late or over budget. But, if upper management understands the expected benefits of the project and is committed to it, your project has a better chance of continued funding.
Solution: Make sure your project has a "champion" in upper management ranks. This champion is one who sees the need for the project, has at least a minimal understanding of the technology involved in the project, and is convinced of the value of the project to the organization. (Helpful hint: If you cannot find anyone in upper management who is willing to champion the project, you may want to take a closer look at it and see if it really will bring significant benefits to your organization.)
Technical Decisions Made by the Wrong People
It is unlikely that a CEO would allow the head of the IT department to decide where to build a new factory, but corporate executives often read an article about some new technology or process and ask the IT department when it will be implemented. Although it is certainly appropriate for management to ask IT to investigate a new technology, it is foolish to insist that a company adopt a technology simply because a major competitor or some other large company is using it. Nevertheless, IT departments are often pressured to install inappropriate technology because upper management is unaware of crucial technical details.
Solution: Take the time to educate upper management about new technologies, especially when asked to implement one or more of them. Evaluate the technology and report your findings to the person who made the request. If they insist on having you implement inappropriate technology, find another member of senior management to take your side on the issue or suggest that the requestor fund the project from his or her own departmental budget.
Delays in Completion
Missed ship dates are the plague of the packaged software industry, with delays of six months to a year common. Thus, it should be no surprise that internal IT projects often suffer from the same delays. Delays are most often caused by changes to the feature set, changes to the underlying architecture of the system, and poor estimating at the start of the project. They can also be caused when a key employee leaves the company or is assigned to a new project, and a new employee must fill the gap and spend extra time getting up to speed on the project.
Solution: There are several inexpensive project management software packages on the market. Choose one, learn how to use it, and carefully monitor progress on your projects. As interim deadlines are missed, immediately find out what is going wrong and take the appropriate steps to correct the problems. Remove any obstacles that are impeding the project, allocate more staff members to it, or change the assignments of your existing staff. But be careful not to cause additional problems as you modify the project timetable, or you might cause even bigger problems later in the project. (It is important to act quickly. The longer a project proceeds with delays, the harder it is to get back on schedule.)
Cost Overruns
Significant IT projects are expensive to implement. They can easily run over budget, especially on projects that last for extended periods (12 months or more). In the USA, experienced IT personnel are commanding larger salaries and more extensive perks (such as cash hiring bonuses) every year. This environment leads many of the most talented employees to change jobs more often, and higher turnover rates lead to increased costs (for training, recruitment, and higher, more competitive salaries). Delays in completion also lead to cost overruns.
Solution: Cost overruns can be minimized or even eliminated with good designs and strong project management skills. Reusing standard components, whether commercial products or items created by your staff, can also dramatically reduce costs and minimize the risk of cost overruns. Reducing staff turnover is another key, whether through excellent leadership and management practices, competitive compensation packages, career advancement opportunities, etc.
Insufficient Testing
Once the IT project has been completed, it must be tested thoroughly before being deployed. That way, the defects can be resolved before the company begins relying on the project for production work. However, IT staff often scrimp on testing. It is expensive to create a proper test plan, and it is expensive and time consuming to conduct the testing. It also takes time, and if the project is near (or past) its due date, the IT staff will be under a lot of pressure to deliver. Thorough testing is usually one of the first casualties of deadline or budget pressures.
When an IT solution is not tested properly, it will usually have serious defects or will not perform adequately when placed under a full operating load. These problems will wreak havoc in a production environment, often bringing work to a complete halt. The problems must then be fixed by the IT staff. In some rare cases, the performance of new software or a new network architecture will be so poor that it ends up being less expensive to discard the entire project and start from scratch than to repair the defects.
Solution: Thorough testing should be built into every project budget. Whether you use standard testing tools, design your own, or do it by hand, never skimp on the testing budget, and never try to make up for delays by reducing the amount of testing you do.
Conclusion
All too many Information Technology projects never come to fruition. Follow these simple guidelines to minimize the risk, and you will dramatically improve the chances of completing your next project, and doing it on time and under budget.
©2000 Information Management Systems, Inc. All rights reserved. Send questions and suggestions to webmaster@infoms.com.