Once you realize that a project is in trouble, the next question is what to do about it.

Sometimes a project can be saved; sometimes it can't. In general, the earlier you spot trouble, the easier it is to do something positive about it—anything from making minor adjustments to scope or schedule to killing the thing outright.

"Early warning gives you the opportunity to make small course corrections early," says Raj Kapur, executive vice president for the Center for Project Management, a project management consultancy and educational organization headquartered in San Ramon, Calif. "The earlier in the lifecycle the issues are found, the easier the team is going to take it."

The first step in fixing a project is getting everyone to admit the project has a problem. If you've spotted the trouble early enough, this recognition can be merely a psychological hurdle. The normal human response is to make explanations and excuses as things start to slip. Often managers and team members don't want to admit that things are going sideways until they get really bad. Or, alternatively, everyone is so busy doing his or her own job that they haven't noticed yet.

"Often, team members are disconnected," Kapur points out. "They're doing their work head down and so focused on doing their part, they really don't get the big picture."

One big help is honesty. "Give your project team the respect of the bad news," Kapur advises. "Don't play games with the team. Especially the seasoned professional will see the writing on the wall."

The next step is to figure out what's wrong with the project. This usually involves careful examination and a lot of hard, close listening. The question always has to be, "How do we fix this," not, "Who is to blame?"

"It's important to avoid blame," says Frank Gianic, president of PMO, an Austin, Texas, consultancy that specializes in rescuing troubled projects. "One of my first speeches when I come in to a project is that we're all starting fresh. Let's not get focused on who did what in the past; focus on the next 30 to 90 to 120 days."

Gianic says it's also important to provide some fast, if minor, successes to help rebuild team morale. "Identify some quick wins to help the team feel the project is getting back on track," he advises. "Typically, you can find a critical path item and work through getting that one thing done. That's critical. Before the team can start to develop some optimism you have to have that kind of early quick win."