Melt Down the Iron Triangle

X'ed out Triple ConstraintSuccess rates for projects as judged by the “Triple Constraint” of scope/quality, budget, and cost, are miserably low. Less than half of any type of project succeeds on just two of those metrics, according to decades of surveys including the most recent from the Project Management Institute (PMI).[1] So many of the factors in success of any business action are beyond the team’s control, it is quite impossible to predict whether a given project will hit even one of those metrics.

This raises what seems like an obvious question: Why use the Triple Constraint (3C) as the standard for success when it is impossible to predictably achieve? PMI doesn’t! Indeed, the terms “Triple Constraint” and its harsher cousin “Iron Triangle” do not appear in PMI’s Project Management Body of Knowledge (PMBOK). Instead it says that at the project level, “Success is measured by product and project quality, timeliness, budget compliance, and degree of customer satisfaction.”[2] Notice that:

  • Scope is not mentioned.
  • “Timeliness” is not the same thing as matching the original schedule, because the customer will likely judge the time required acceptable if happy with the process and final product.
  • The PMBOK includes a process for “managing changes to the cost baseline” and adds that this “process is performed throughout the project,” indicating the term “budget compliance” need not refer to the initial budget.

Critically, there are similar processes for the other 3C factors, meaning PMI considers all three changeable. This is another reason those factors are worthless as success metrics. In a sense, every project should hit the 3C factors every time, because the project plan is supposed to be updated (“rebaselined”) when new information comes in!

Nonetheless, most bureaucracies have trapped themselves within a myth that initial Triple Constraint numbers in a project plan are inalterable promises that can (and must) be met every time. Executives know full well how impossible this was when they were further down the chain, at least without bullying workers, and the games they still have to play to “make their numbers.” I witnessed a manufacturer delaying shipments by a week, knowingly disappointing customers, to keep from earning too much money in a quarter. Many executives go along with the lie that estimates represent truth, rather than working together to change the culture surrounding that lie.

Public companies will argue, “The stock market demands we meet projections.” Thus 3C rigidity is justified because you have to know which scope is going to be delivered within a given period, and how much it will cost, to create those projections.

Except the market doesn’t demand them. The 2003 book Beyond Budgeting from Harvard Business School proves it possible for corporations to thrive without trying to hit projections.[3] And there’s no legal requirement for them. Securities lawyer and SEC expert Brian Lane said at a conference, “it’s a myth that projections are required… You do have to talk about known uncertainties and how they could impact the future, but you don’t have to make projections.’”[4]

More fundamentally, research from various sources like the American Association of Independent Investors, European Central Bank, and McKinsey & Company found the primary driver of stock value is long-term earnings and returns (see “Annual Budget Games”). Prices may take a short-term hit when you are far off from company or analyst projections. But you can limit those impacts by not making projections yourself and being more transparent with analysts, so they can adjust theirs in real time. That is, provide bad (or good) news before it becomes a bad surprise.

Beyond Budgeting and the personal experience of anyone who has been involved in the standard budgeting process can tell you it takes absurd amounts of money to perform and manage, in exchange for little long-term value. A board of directors acting on purely rational grounds would demand the CEO stop wasting company time on fixed financial predictions versus, for example, rolling annual budgets that are updated monthly as new figures arrive.

Another defense of 3C estimates is that “customers want them.” Do they really? Think about things you ordered online. Are you more irritated by those which arrive late but right, or by those arriving on time but broken, or missing features you expected? As I detail at “What Customers Really Want,” strong evidence suggests that customers ultimately want to be satisfied with what they get after they get it. I’ve read numerous studies in which 3C standards fell well down the list of factors in customer ratings of success.[5] PMI recognizes this in its PMBOK definition of “Program” success: “the program’s ability to deliver its intended benefits to an organization, and by the program’s efficiency and effectiveness in delivering those benefits.”

Your company or agency may be heavily dependent on customers that require 3C estimates as part of a Request for Proposal (RFP) process. This is why I begrudgingly include a Lean estimation process on my Full Stack Scrum™ site, to come up with those numbers as cheaply as possible. Companies and governments issuing RFPs create a ripple effect of pain as vendors lowball the estimates (lie); vendor leaders beat on employees to meet the unrealistic numbers; customer contacts beat on the vendors; those contacts get beat on by their leaders; and so forth. There are alternative ways of setting up those contracts to ensure the end users get what they need as cost-effectively as possible. (See “Agile Contracts” for a detailed example.)

The rational solution is to eliminate use of the Iron Triangle factors as performance standards. Measure them if you must, but as tools for managing resources, not as unfair standards for success. Given their low predictability rates and the degree to which achievement of initial estimates is beyond the control of any group of people (much less one project manager), tying performance appraisals and compensation to them borders on being unethical.

Since the Agile Manifesto (which I apply to all types of work, not just software) implies, and management research shows, that the ultimate path to financial success is to please the customer, customer satisfaction is the best performance standard. Close collaboration and iterative development as called for in the Manifesto are the easiest ways to achieve it. If the customer is closely involved with the planning and evaluates deliverables regularly—be they updated products, or design or construction plans with newly completed sections—the customer will:

  • Know the final product will meet their needs when finished, because the output or its plans are refined as their needs change.
  • Predict for themselves how their new requests will impact costs and schedule, because they will understand the interaction of user stories, backlog rank-ordering, and velocity.
  • See that you are working as fast as humanly possible within the bounds of a sustainable pace.
  • Understand “why it is taking so long” or “costing so much,” instead of wondering or having to ask.

These, in turn, nearly guarantee high customer satisfaction, because the customer realizes they have primary control over the project—and thus their own satisfaction!


Please share this post at the bottom of the page.

[1] PMI (2018), “Research Highlights by Industry and Region,” Pulse of the Profession (Project Management Institute, 2018), Although the report does not provide a figure for meeting more than one factor in a given project, the success percentages for the individual factors range from 52% to 57%. Mathematically the figure for meeting any two factors must be below 50%, and for all three even lower.

[2] Project Management Institute (2017), A Guide to the Project Management Body of Knowledge (PMBOK® Guide)—Sixth Edition. Project Management Institute: Newton Square, PA.

[3] Hope, J., and R. Fraser (2003), Beyond Budgeting: How Managers can Break Free from the Annual Performance Trap. Harvard Business School Press: Boston.

[4] Ken Tysiac, “Nine Tips for Effective MD&A Reporting,” Journal of Accountancy, December 4, 2012,

[5] Examples from consulting and process change projects include: Guy G. Gable, “A Multidimensional Model of Client Success When Engaging External Consultants,” Management Science 42, no. 8 (August 1996): 1175–98,; Alan Simon and Vanya Kumar, “Clients’ Views on Strategic Capabilities Which Lead to Management Consulting Success,” Management Decision 39, no. 5 (June 2001): 362–72,; Dag Näslund, “Lean and Six Sigma – Critical Success Factors Revisited,” ed. Su Mi Dahlgaard Park, International Journal of Quality and Service Sciences 5, no. 1 (March 22, 2013): 86–100,; Matias Bronnenmayer, Bernd W. Wirtz, and Vincent Göttel, “Success Factors of Management Consulting,” Review of Managerial Science 10, no. 1 (January 2016): 1–34,

Tell the world: