Post-Mortems Key to Successful Future Projects
A common misconception about post-mortem reviews is that they are only needed when something goes wrong. Not true, says Joseph Phillips, author of IT Project Management: On Track from Start to Finish. According to Phillips, the post-mortem really starts at the beginning of the project when goals are established. If the goals are met, the post-mortem sets out to prove it; if the goals aren't met, then it tries to discover what went wrong.
At its heart, the post-mortem is an end-to-end review of the project from its conception to its delivery. A thorough post-mortem will call on the involvement of personnel beyond the developers who made the project work. Issues with procurement must be addressed; finance people must be involved to keep the budget -- and the budget of future projects -- in line with the product.
When it comes to spending money, forecasting is done upfront when a project is still in the concept stage. When it's time for the post-mortem, it's time for accounting for the money. At a large organization, it's often a formal audit that accounts for how money was spent. Project managers at such organizations often have autonomy on spending up to $5,000 or $10,000. If spending exceeds this, then the sponsors of the project get involved along the project's lifecycle to review funding. Smaller organizations might have a more informal review.
"The heart of the post-mortem is verification," Phillips says. For most projects, the post-mortem review should be done at final acceptance of the project, as soon as possible to the final sign-off. The reasons for this are simple: members of the team are going to move on to other projects; and the experience of working on the project will be fresh in everyone's mind soon after it's done.
Exactly what goes into a post-mortem varies greatly from organization to organization, but even the smallest of organizations or development groups needs to have a formal process, according to Phillips. A formal process will make it easier to find what went wrong, even in a project that was largely successful.
"More than 'try again next time,' there has to be a formal process," Phillips says. The process itself will call upon what Phillips says are the two greatest traits of a good project manager -- organization and leadership.
Organization is a natural because project managers have to know what multiple people or groups are doing, as well as coordinating deadlines. When it comes to project management, planning is the key to success.
Good leadership skills are important not only because someone has to be in charge, but especially when things go bad, someone has to lead. It's a proven fact that people work better when they sense someone is in charge. Leadership also comes into play because someone has to keep the post-mortem from getting ugly if the project encountered problems.
When Post-Mortems Go Bad
"It's always easier to go through a post-mortem after a successful project," Phillips says. But a post-mortem is most important after a bad project, otherwise more bad projects and more scary post-mortems are sure to follow.
According to Christopher Avery, author of Teamwork is an Individual Skill: Getting Your Work Done When Sharing Responsibility and a speaker and advisor to companies, the clear intention of a post-mortem is threefold: learn, correct, and improve. If you venture beyond these three goals, you're asking for a trouble, namely a giant blame game.
Avery says time should be made at the beginning of any post-mortem meeting to make clear what the goals are. The meeting should then be run with such process that it is kept within the boundaries. The hard part is holding close tolerance so things do not get out of hand, but at the same time making it clear that there is no punishment for telling the truth.
"Our primary mode of learning is making mistakes and learning," Avery says. "You have to base [the meeting] on acknowledging mistakes and correcting them."
The entire post-mortem process should not only be open to everyone who was involved with the project, but should be conducted in front of the group. In those cases where some sort of disciplinary action is required, that should, of course, be done one-on-one.
"The most important thing," Avery says, "is are we intending to make decisions based on reality, or are we intending to cover our butts and find scapegoats?"
A good post-mortem after a bad project relies requires honesty and a willingness to improve, both of which can be inspired by a good leader. Some organizations that have well-defined methodologies might also have something in place to ensure that any action that needs to be taken as a result of a post-mortem is done. Together, strong leadership and process that ensure improvements are made will increase the likelihood of successful projects.