It is a horrible sight to watch a project fail. There are few things as bad. Watching a project fail regardless of the reason is almost like sitting in a room with a "Dementor" from Harry Potter. It will literally suck all of the life and joy out of the room.

Nearly every project that I have seen fail has failed because of political challenges or management challenges. Sometimes there are technical challenges that bring a project to its knees, but usually projects fail for less technical reasons.

Here a few observations about projects failing for political reasons.

• Both the client and the consultants have to be committed to seeing the project succeed. Put simply, you cannot solve a problem when the primary stake holders do not truly want it solved. This could come from a consultant being more interested in extended the engagement. It could come from a client being afraid of what will happen to them once the problem is solved. It could come from disenfranchised stake holders. Sometimes a project is beset on all sides. When you find yourself working on a project that has this kind of threat, do all that you can to constrain the disruptive influences of the bad apples. If their influence cannot be constrained, you truly have no choice but to move on to a new project.
• Tough choices have to be made to make a project successful. These choices will affect everyone involved in the project. These choices may involve users not getting a change request through that they want. Developers may not get to use the tools that they want. Everyone may have to put in more hours that they originally planned. Steps may be skipped. Compromises will be made, but if everyone stays committed to the end goal, you can still be successful. If individuals start feeling disgruntled or resentful of the compromises reached, the project can easily be derailed. When everyone is not working towards a common goal, it is like driving with one foot on the break and one foot on the accelerator. Not only will you not get to where you are planning, you will also damage the car and possibly the passengers as well.  
• It is important to always keep the end result in mind. Regardless of the development methodology being followed, the end goal is not comprehensive documentation. In all cases, it is working software. Comprehensive documentation is nice but useless if the software doesn't work.  
• You can never get so distracted by the next goal that you fail to meet the current goal. Most projects are ultimately marathons. This means that the pace must be sustainable. Regardless of the temptations, you cannot burn the team alive. Processes will fail. Technology will get outdated. Requirements will change, but your people will adapt and learn and grow. If everyone on the team from the most senior analyst to the most junior recruit trusts and respects each other, there is no challenge that they cannot overcome. When everyone involved faces challenges with the attitude "This is my project and I will not let it fail" "You are my teammate and I will not let you fail", you will in fact not fail. When you find a team that embraces this attitude, protect it at all cost.

Edward Yourdon wrote a book called Death March. In it, he included a graph for categorizing Death March project types based on the Happiness of the Team and the Chances of Success.

Chances are we have all worked on Death March projects. We will all most likely work on more Death March projects in the future. To a certain extent, they seem to be inevitable, but they should never be suicide or ugly. Ideally, they can all be "Mission Impossible" where everyone works hard, has fun, and knows that there is good chance that they will succeed.

If you are ever lucky enough to work on such a project, you will know that sense of pride that comes from the eventual success. You will recognize a profound bond with the team that you worked with. Chances are it will change your life or at least your outlook on life.

If you have not already read this book, get a copy and study it closely. It will help you survive and make the most out of your next Death March project.