Do you want to learn how Nobel Prize winning research can help you deliver your projects on time and on budget, consistently? If so, this article is for you.
Imagine this scenario - you take on a new project, you know precisely how to plan for the project to achieve a particular profit and margin goal, and you execute the project just as planned. The outcome? A happy team and a happy client - everybody wins! The problem is that this imagined scenario rarely is reality. In fact, a McKinsey study on over 5,000 organizations found that half of all large IT projects massively blow their budgets. On average, 45% run over budget and 7% run over time, while delivering 56% less value than predicted. And this is across industries; no one seems to have it right.
The question is - why is this our reality? According to economist Daniel Kahneman, who won a Nobel Prize for his work on this topic, projects are doomed from the start because of something called the planning fallacy. The planning fallacy - which tends to lead to poorly planned projects that are sure to go over budget, exhibit scope creep, and extend deadlines - is a result of human overconfidence and the tendency to neglect distributional data about the outcomes of similar projects. In simpler terms, humans are often bad at planning projects because they fail to consider the outcomes of past projects when planning for a new project.
It's a good thing then that Kahneman recognized the planning fallacy and introduced a solution to this prevalent problem. He named that solution Reference Class Forecasting. Reference Class Forecasting (RCF) is based on the idea that the person(s) doing the project planning should ‘make every effort to frame the forecasting problem so as to facilitate the utilization of all the distributional information that is available to the expert’. Again, in simpler terms, project planners should utilize information about prior similar projects (the ‘reference class’) to forecast the demands of the current project.
Kahneman describes the following three steps to become a reference class forecaster / project planner:
- Select a reference class of past, similar projects. An example might be all past website builds that your company has performed across clients.
- Establish a distribution from projects in the reference class. One project may have finished with 200 hours of work while another took 300. This spread is the distribution within your reference class. In this example we’ve used time, but it could be another metric like cost, client satisfaction, etc.
- Compare the specific project with the reference class distribution, in order to establish the most likely outcome for the specific project. Use your human knowledge and experience to determine how your current project differs from your reference class. This, Kahneman calls an ‘intuitive’ estimation. Then, interpolate between the reference class and your intuitive estimation to come to an accurate estimation for this project.
The beauty of Kahneman’s work on RCF is the insight that the best solution to the forecasting problem is to take into account ‘distributional’ data about prior projects. The reason this is so powerful is because, if you consider multiple phases of a project, the probability of any individual phase running awry is relatively low. However, when you link together multiple phases of a project with multiple stakeholders and, often times, elements that are out of your control (last minute client demands, project members getting sick, miscommunications, etc.), the joint probability that at least one of these elements will go awry is rather high. Thus, Kahneman’s insight that distributional data from prior projects should account for these unexpected setbacks is a sound one, and powerful enough of an idea to help him win the Nobel Prize.
Of course, in order for RCF to work for you, you must have data about past projects that you can use as a reference class for future work. However, gathering such information and storing it in a way that’s easily usable is one of the more challenging parts of successfully implementing RCF, and one of the reasons that RCF isn’t widely used by all project based organizations.
That’s precisely where Allocate can help. Having a deep understanding of the work that Kahneman did on RCF to win the Nobel Prize, we’ve built our project planning and time tracking solution in a way that allows you to become a reference class forecaster, likely without even knowing it.
What does this mean? Allocate was built to facilitate the project delivery life cycle, and a core component of that is reference class forecasting. Let’s briefly explain.
The project delivery life cycle includes the following elements:
- Taking on a new project
- Creating a sound project estimation, driven by reference classes that are easily stored in Allocate from prior projects
- Staffing that project with the right people who are available at the right time
- Launching the project, and tracking progress against that project on a monetary (how much time has been logged to this project - our AI-powered solution helps you get more accurate, more real-time information), temporal (how much time is left until delivery), and milestone basis (which core tasks have been accomplished).
- Analyzing data about the project progress, and making any course corrections or other changes necessary
- Delivering the project
- Analyzing what happened during the project, and using the learnings as a reference class for future projects
… and the cycle continues!
The cycle graph above explains this process in more simplified terms.
In summary, RCF is a powerful, Nobel Prize winning theory for avoiding the cost and time overruns that affect the majority of projects, across industries. The McKinsey study empirically demonstrates this, and chances are that your own experiences also support the findings. If you’re interested in harnessing the power of RCF to deliver more projects on time and on budget, while achieving your desired profit and margin goals, then schedule a demo of Allocate to see how we can help.