If you are a business leader responsible for delivery, the scenario of slipping dates is not new. However, it is ugly, including the conversations and post mortem that follows. When a project does not get shipped on time, or a customer issue or failure occurs, it is very common to do a post-mortem and identify lessons. And, post-mortem is very important.

In 2007, Harvard Business Review published an important article around premortem. The basic idea of premortem is this. When you start a major project, you essentially indulge all the key leaders in a thought exercise centered around a futuristic scenario of your project failing. You start with the premise that the project is a major failure and have everyone write a reason for why that happened. This is purely intellectual and because this is not a desired outcome, you end up with a diverse set of perspectives, some of which might be potential blindspots. Then, you work through a project plan that takes all the reasons into account. By doing this, you are minimizing the surprises that might sneak and derail the project at a later stage.

This is important. There has been enough studies [123] about our tendencies to under estimate the risks. Even experienced leaders are not immune to quality, personnel headaches and project deadlines slipping.

In a particularly interesting study, researchers created experiments to understand how experience based learning works in seasoned managers and they identified three aspects of why projects fail. The themes that emerged are -

  1. Time lags between causes and effects - People find very hard to link causes and effects, let alone learn from them. Because cause and effect has a time lag, people show the tendency to repeat the mistakes. This includes hiring required talent and assimilation of information and complexity critical to making a project successful.
  2. Fallible estimates - Estimates around team productivity and work required were below par.
  3. Initial goal bias - It is important for managers to adjust and course correct with new information about project health and dependencies. The study showed that initial goal bias, which is a set of goals and assumptions regarding cost, schedules and time, does not calibrate as required.

The study derives an important corollary.

Despite their experiences with complex projects, the veteran managers do not meaningfully improve the mental models they have formed in simpler contexts.

Having a growing mental model is critical to projects becoming successful. I believe this is where premortem helps. The nature of premortem allows for a thought exercise where the effect is put before cause which allows exploring the reasons for why a project might fail before it does. This will create an environment where mental models can be improved without the unfavorable time lags, which hinder learning from failures.