IT project bleeding cash? Consider pulling the plug: a consultant offers some proven signs for determining if a technology project is foundering and might be better off being canceled, even at the expense of the outlays already made.

It's that time of year, when you are working on next year's budgets. You finally get to the IT department, and you don't know what to do, because there's one project that is over budget--way over. Should the company continue to fund it, even though you suspect it is simply bleeding cash? And, if you decide to continue funding it, how can you be sure your company has the ability to contain it? Or, should you simply pull the plug and stop the bleeding?

This is far from an unusual choice. According to a recent survey by The Standish Group, 51 percent of IT projects are challenged and 15 percent fail outright. The last thing you want to tell the CEO (especially if your company is publicly traded) is, "We are going to take a multi-million dollar charge because of a failed IT project."

To determine whether or not you should pull the plug, you must first see if the project is out of control (or is about to get that way). There are several signs to look for.

The most obvious sign is that people keep coming back for more money. You need to ascertain if this is an anomaly or a trend. The project manager should have assigned a dollar value to each of the project's deliverables. How often do the actual dollars spent on associated deliverables exceed the planned budget? This should answer that anomaly or trend question--especially if the "overage" was not approved.

Another obvious sign is that deadlines are being missed. Again, you should look at the frequency to determine if missed deadlines are another anomaly or a trend.

The next two signs are actually related to things that should have been in place before the IT project started. The first is the project charter, also called the statement of work. This defining document of an IT project indicates things such as how the deliverables are defined and who has acceptance authority, roles and responsibilities for the extended project team and the scope of the work to be done; it should have been agreed on before the project was started. Ask the CIO for a copy. If there's no project charter, it is a sign that you probably have or will have a runaway project.

Next, ask to see the project plan and find out when was the last time it was updated. (It should be updated weekly.) If there is no plan at all, or the plan has not been updated, you probably have (or will have) a runaway project.

Also, tasks with large chunks of hours associated with them in the project plan represent a major danger sign. It is difficult to measure progress on tasks that take more than 80 hours. And hour 180 of a 200-hour task is not the time to find out that the deadline is going to be missed.

The project charter and project plan are key because without them, it's difficult to tell just how bad your situation is and who to hold accountable.

A final sign that you may have a project that is bleeding cash is people not knowing what tasks they are supposed to be working on. If they are not working according to the project plan, it is a sign that the project is not being managed properly.

It was immediately obvious that one client's project was bleeding cash when the project manager, programmers and business analysts could not show which tasks in the project plan they were working on. In fact, several programmers stated outright, "My tasks are not in the plan." Clearly, immediate changes were needed.

Let's say you've seen some or all of the above signs, and have determined that you have a runaway project that is bleeding cash. You next have to determine and validate the project's current ROI.

In most cases, someone had to do a business case to get the project approved and funded. What is the project's ROI based on? Is the project supposed to affect the top line, the bottom line or both? If the project is bleeding cash, the previous ROI projections have changed and must be recalculated. After doing this, ask yourself if the ROI still makes sense and if the business benefits make the project worth continuing.

Now you've arrived at the first decision point to determine if you should to pull the plug on the project or get it under control. Next, you have to put the project back in the IT-governance hopper. Compare the value of the project to other IT alternatives and initiatives. If you have good IT-governance, you may determine that you can get a better and faster return with another project. You are now at another decision point.

Pulling the plug, however, isn't an easy thing to do--and it often isn't the right thing. An example is implementing a new enterprise resource planning (ERP) system; you may have to do it because the legacy applications can't give you the needed insight to run the company. However, one company implemented an ERP system for $30 million dollars, determined the new system didn't meet its needs, pulled the plug on it and went back to the previous system. (You've got to believe that the company could have determined it needed to pull the plug long before it spent $30 million dollars!)

With a runaway project that must get done, it is even more important that you contain and control it. The project can't go on as is; it will not just fix itself. There are several steps you must take to get a project that is bleeding cash back on course:

* Admit you have a problem. This is difficult because many people say, "I'm this close. I just need one more month." However, that last month never comes.

* Pause the project. By continuing, you are burning more hours and cost against the project and you really don't know where you are going. Pausing the project gives you an opportunity to regroup and create a new plan. It is the hardest and the scariest thing to do, but if you don't stop, the results down the line will be even scarier.

* Conduct a project audit. The purpose of the project audit is to find the root cause of why things are going wrong. Usually, the project manager is relatively inexperienced and there isn't a support system in place to help him or her overcome this inexperience.

* Assess the effort to complete the project. Find out what it will realistically take to finish the project and how much it is going to cost.

* Restart the project. Now that you have a new project plan and new estimates, re-launch the project.

Unfortunately, because of the complexity of IT projects, many financial executives do not exercise the same due diligence with them as they do with non-IT projects. However, by knowing the early warning signs of an IT project that is bleeding cash, you can contain and control it, or at least pull the plug before more money goes down the drain.

Joseph J. Zucchero is Executive Vice President and Director, Project Turnaround Practice, in the Tampa office of The Casey Group, headquartered in Parsippany, N.J. Zucchero has years of experience dealing with program and project management and the execution of highly sophisticated IT contracts. He can be reached at 813.282.6977.