Why Killing Failing Projects Is So Difficult

Failure1In organizations where multiple projects are being managed simultaneously, inevitably there is a sense among management as well as project teams that some projects should be postponed or killed outright and the resources re-allocated to higher value activities. Organizations regularly continue to invest in projects even though they know they are failing. Maybe there are technical hurdles, or the competitive situation has changed. In theory, the phased-development processes aid in making those decisions at the go/no-go gates for each phase. The reality in most organizations is much different. While senior management knows that some projects need to be shelved, why is there such resistance? This article addresses some of these issues. 

First, we need to consider how we make decisions. It is important to understand what drives our behaviors in order to make conscious decisions about how we might change our approach. A colleague, Paul O’Connor from the Adept Group, recently published an article entitled “Irrational Thinking and Clueless Portfolio Management” (1) and described his take on the two ways we make decisions based on microeconomic theory. We make decisions either rationally, using facts and data only, or when that becomes too complex, we fall back into decisions rooted in our emotions. Or as Paul likes to say, “…our inner Homer Simpson can take over.” As anyone who is involved in managing innovation knows, new product development (NPD) decisions are fraught with emotions, so we have to consider how our emotions impact our decisions.

Other research labels these processes “System 1” and “System 2” cognitive strategies (2). System 1 refers to our automatic, almost effortless decision-making process that is based on intuition and emotions, or our “gut feel”. System 2 processes in contrast are based on logic and data, and are conscious and explicit processes. While System 1 is very fast, System 2 takes much more time though it may lead to more effective decisions. The problem is that when faced with the time and resource constraints so common in business today, we often fall back into System 1 processes exclusively. For some senior managers, their career may be based on making successful System 1 decisions and have the reputation as a “visionary” leader. Of course, one miscalculation can send a career in an entirely different direction.

In addressing our potential over-reliance on System1 thinking, first we have to improve the quality of the data that we have regarding the portfolio and be able to effectively compare and contrast projects. The Adept Group’s Portview™ is one such tool. While this tool allows for plotting project variables in any number of ways in a graphical format, the real value is the discussions that ensue as a result of collecting the data. Suddenly, the quality of the data becomes very important. It is very common that when a company collects high quality data and compares for instance some measure of cumulative project value vs. a measure of cumulative resources required to achieve that value for the entire project portfolio, they find that a very high percent of the cumulative value could be achieved with much less percentage of the cumulative resources. This is illustrated in the figure below.

Portfolio Entropy

Looking at a portfolio in this way encourages honest discussion, or should, about the projects that make up that last 10% or so of cumulative value, and the fact that the organization is expending increasing amounts of resources for lower value. There may in fact be good strategic reasons, but the decisions to continue on with those projects should be a conscious one. So often, there is no discussion about the projects in the portfolio and simply no decision made. Unfortunately, no decision is actually a decision if resources continue to be expended. The concept of “disorder” in the portfolio is an apt analogy (3). All systems, including a portfolio of projects, naturally tend to a higher state of disorder. For instance, new projects are added without any regard to existing projects and the resource constraints. The only way to control the situation is to expend energy, management energy in this case, to reduce the chaos and establish a higher level of order.

While improving the quality of the data and using tools like PortView to encourage a discussion are important ways to increase System 2 thinking, it is not the total story. For instance, in the research previously cited (2), improving the quality of data particularly in terms of projects that are failing reduces some of the factors that lead to poor decision making, but it does not eliminate them all. The role of emotion in our decision making has to be considered. There is a concept from this same research called “anticipated regret”, which is comprised of “anticipated keep regret” and “anticipated drop regret”, and how they pull decision makers in two different directions.

As the author’s describe these two emotional factors:

“Anticipated keep regret is felt when decision-makers allow a project to continue but anticipate looking back from the future and deciding this was a mistake. Contrarily, anticipated drop regret is felt when decision makers decide to discontinue a project yet worry about viewing this decision as a mistake in the future when a competitor introduces a similar product, for example.”

EmotionsWhat is interesting about this research is that the author’s found that the effects of these two emotions are not equal and differ in direction and magnitude in how they influence decision making. Anticipated keep regret increases as a failing project progresses, and as more negative information is received that the odds of success are decreasing. This is rational, System 2 thinking, though that does not mean a project will be killed. There are still many emotional reasons why the organization will be biased against killing a project even if anticipated keep regret increases as a failing project progresses. On the other hand, they found that anticipated drop regret remains essentially constant even in the face of mounting evidence that the project is failing. As the author’s opine, it seems that:

“Decision makers seem to cling to slim chances; they anticipate regret and abandon rational thinking and use emotional (System1) thinking. They keep the option to complete the project open and thereby keep a chance of success, however small, because terminating it guarantees failure…Dropping an NPD project appears to be tough emotionally because decision-makers continue to anticipate feeling regret for their prior investment and lost opportunities.”

There are of course powerful forces that drive our emotions and System 1 thinking, both from the standpoint of senior management and project teams. For instance, despite the fact that sunk costs cannot be retrieved, we allow them to influence our decision making. Another factor is that managers do not get rewarded for unfinished projects. There is a natural bias, especially for project managers, not to recommend killing a project as that may not be viewed as a way to enhance one’s career. This can apply to senior management as well if they proposed and supported a project. How close a project is to completion also influences our decision making: we may be averse to killing projects the closer they get to the finish line. There are also social structure factors. A project leader will be influenced by his project team members who may view their livelihood and continued employment hinging on decisions about continuing or killing a failing project. How senior management rewards or punishes “failure” and how open they are to receiving “bad news” related to a project may be important. If project leaders sense that they will be punished for admitting that a project is in trouble based on System 2 thinking, then an honest exchange of information may not occur. If the senior manager does not want to hear bad news and have an honest discussion about problems and solutions, and simply insists that the project cannot fail as defined and the team “just needs to work harder”, that also propagates emotional and sometimes irrational System 1 thinking.

In the final analysis, it will be up to the senior manager who is responsible for the P&L in the business to make sure that he/she insists on improving the System 2 processes using effective tools for evaluating the project portfolio and improving data quality. While “gut feel” should not be eliminated from the decision making process, it should not be the only methodology. On the other hand, the senior manager must recognize the role that emotions or System 1 thinking plays in making decisions about failing projects and portfolio entropy in general. This applies not only to the organization and project teams, but to the senior manager him/herself. It will likely rest with the senior manager to make sure that emotions are held in check, and that the culture supports a healthy balance between System 1 and 2 decision making processes. Ultimately the senior manager will have to stand alone and make the hard decision to kill a project(s) tracking to failure or consuming too many resources for the economic value and bring a project portfolio in chaos back to a degree of order.

What is your experience in your organization with killing failing projects? How averse is the organization to doing that? Do you use tools to help evaluate your portfolio? What are they and how have they worked? What other emotional drivers have you found that influence decision-making on failing projects?


(1)O’Connor, Paul. 2014. Irrational Thinking and Clueless Portfolio Management. Adept Group. June 2, 2014.

(2)Sarangee, Kumar, Schmidt, Jeffrey B., and Wallman, Jeffrey P. 2013. Clinging to Slim Chances: The Dynamics of Anticipating Regret When Developing New Products. Journal of Product Innovation Management  30(5):980-993.

(3)See Assessing Disorder in a New Product Development (NPD) Process

New Product Visions is a consulting company that helps organizations improve the effectiveness of their new product development processes. We specialize in small to mid-sized companies that manufacture highly engineered products. Contact us today about how we might help you!

Specialties: NPD consultants, new product development consulting, developing new products, new product development seminars, small business consulting, new product development expert, product development process, new product development strategies, integrating NPD for mergers & acquisitions, organizing for innovation, management role in NPD, project risk analysis, innovation management



1 thought on “Why Killing Failing Projects Is So Difficult

  1. Great insight. I would add one more tip, taken from the startup world, use MVPs (Minimally Viable Product) to validate the market early. Our research shows that most NPD fail because there is not a good product fit with the market. Short and cheap, the MVP can confirm or invalidate a product idea before spending large budgets. MVP are also easy to outsource, so no need to divert internal resources on a project that may never see the light of day.

    Keep up the good work
    Steve Owens

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.