Project management & Quality Assurance is really just about two things.

1. The organisation should be aware of itself, its projects, processes, roles and responsibilities. 
2. The organisation should be able to store, share and accumulate knowledge.

Software projects add another dimension to the traditional project development process. The interim stages of the product are not neccessarily visible, so it is hard to track progress, feel momentum or see if the project has taken the right direction. Therefore more precise info must be produced by the projects in order to enable management to make sound decisions.

Information must be structured, stored, accumulated and shared in order to be useful. Storing is not enough. Structuring is needed in order to be able to find the information again. This is the art of knowledge work: The conversion of chaotic information into structured and useful knowledge. It is a prerequisite for success in any organisation, as massive transfer of non-written complex knowledge is bound to fail.

A few simple checkpoints before setting sails

Planning a project is like preparing for a journey.

You need to identify your location and your destination to set a course. 
You need a transport with a crew and a skipper to steer it.
You must have funding, resources and an ETA to plan for.

Also, if you do not know the destination, how will you recognise it, when you see it?
The skipper sets the course. He identifies the waypoints and possibly where to get fresh supplies. 
The skipper then discusses the probability of success and the willingness to fail or delay with the stakeholders. 

In uncharted waters an experienced skipper has obvious but hard to measure advantages such as experience.

The three parameters

There are three paremeters involved... resources, destination and time and they are tied together in the 1st law of project management: Destination / Resources = Time; Obvioysly only two of these can be chosen arbitrarily at any given time. What is less obvious is that it only works within certain boundaries and in a non linear fashion. Doubling resources might double the speed on a long term project, but it will slow things down on a short term project.

The learning organisation

It is definately a good thing to stop and think about what went wrong. In fact, this is the basis for experience. Unfortunately, the real world often moves forward at an ever increasing pace, so many failed projects are left unanalysed, lesson unlearned.