In these difficult times, lots of projects are getting cancelled, postponed or mothballed. Although these are perfectly normal occurrences in IT, they seem more frequent, swift and stinging now. When a project is killed, we like to think that its fate is entirely due to external forces – the swirling, uncontrollable winds of the economic hurricane happening outside. It's not that we're doing anything wrong, we reason; it's just a response to the crisis. However, we know better. Many of those projects weren't selected for cancellation just because of sudden shifts in priorities. Some of them were cancelled because there were problems. They were judged unlikely to ever be completed. Or they were expected to exceed time or budget constraints, or to fail to offer sufficient business value even if they did deliver a product. So in these austere times, it's more important than ever to recognise project problems early. Thus you can correct them or at least cancel the project before wasting too many resources. Yes, we all know that something's wrong with the project when we blow a budget or miss a major deadline, but how can we know before it's too late to do something about it? Here are five early warning signs that your project is in trouble: • Management direction is inconsistent or missing. If project leadership has gone AWOL, chances are that things are starting to go in a bad direction. Or, even worse, if the directives you get from management (or feel compelled to give if you are management) change frequently, there's a problem. If a project either lacks direction or can't maintain a reasonably consistent course, it's unlikely to get to any desirable destination. • Project management and business management seem disconnected. Even if a project does get consistent direction, if that direction seems to be at odds with business management's desires, there's a problem brewing. In political battles between IT and business management, business management usually wins, even if it takes a while. I don't hear too many stories about the great political triumphs of IT managers over their users or clients. • The team lacks a commitment to clearly articulated and commonly understood goals. Every project has a goal or two. They may be clearly stated or only vaguely discussed, but it's rare for any business to shell out lots of money for something that genuinely has no purpose. That said, it's common to presume that the purpose of a project is so obvious as to not be worth articulating. That's unfortunate. It typically leads to misunderstandings and inconsistent presumptions about priorities. Eventually, poor and inconsistent tactical decisions undermine project progress. • Team members don't listen to one another. Even when teams get along personally, team members don't always listen well to one another. This tends to lead to chaos as people fail to coordinate activities and make the compromises necessary to enable projects to move ahead. • The team is in a state of discord. Teams sometimes break into competing camps. These can form around honest-to-goodness differences over project direction. They can also form over petty loyalties and personality clashes. Sometimes teams just descend into chaos, with multiple factions or an every-person-for-himself ethos. The state of discord is destructive to progress. It needs to be rooted out. Sometimes, as a manager, you can engineer a reconciliation. Other times, you need to pick winners and losers. If you notice any of these things happening on your projects, the time to act is now. Don't wait until it becomes obvious that the project is a disaster in the making. By intervening at the first sign of trouble, you may be able to save both projects and careers from the swift executions going on today.