The project was supposed to take six months. Eighteen months later, it finally launched: over budget, under-featured, and late enough that the market had moved on.
This isn’t a story about one project. It’s the story of most projects. Studies consistently show that we underestimate how long things will take, how much they’ll cost, and how many obstacles we’ll encounter. Not occasionally. Systematically.
Psychologists call this the planning fallacy: the tendency to make optimistic predictions about time and resources even when we have evidence from past projects that should tell us otherwise. It affects individuals planning their days, teams planning their quarters, and organizations planning their strategies.
Understanding the planning fallacy doesn’t automatically fix it, but it does suggest approaches that help.
Why we plan optimistically
The planning fallacy isn’t stupidity or inexperience. It persists even among experts who have been wrong before. Several cognitive forces drive it:
We plan for the best case. When envisioning a project, we naturally imagine how it will unfold if everything goes well. Tasks completing on schedule. Dependencies aligning. No surprises. This becomes our baseline estimate, even though “everything goes well” is the least likely scenario.
We underweight base rates. How long did similar projects actually take in the past? We know this information, but we give it less weight than our intuition about this particular project. “This time will be different” feels true even when it rarely is.
We fail to imagine the specifics of delay. It’s easy to acknowledge abstractly that “something will go wrong.” It’s harder to anticipate the specific meeting that won’t get scheduled until next month, the team member who leaves mid-project, the requirement that emerges after development starts. Each delay is individually unlikely; collectively, delays are nearly certain.
We’re motivated to underestimate. Optimistic estimates get projects approved. Optimistic timelines win deals. Realistic estimates feel like sandbagging or lack of ambition. The incentives push toward optimism even when experience suggests pessimism.
We anchor on the plan. Once an initial estimate exists, it becomes an anchor. Subsequent revisions adjust from it incrementally rather than re-estimating from scratch. Even unrealistic plans create gravitational pull.
The evidence is overwhelming
This isn’t speculative psychology. The research base is extensive:
Major IT projects average 45% over budget and 7% over time, while delivering 56% less value than predicted, according to McKinsey research on large-scale technology initiatives.
Mega-projects (infrastructure, construction) come in over budget 90% of the time, with average overruns of 28%, per research by Bent Flyvbjerg.
Even personal estimates for simple tasks show consistent optimism. Students asked to estimate when they’d finish their thesis (both best-case and realistic-case) actually finished closer to their own worst-case estimates.
The pattern holds across domains, experience levels, and cultures. Expertise doesn’t eliminate the bias; experts are simply overconfident in more sophisticated ways.
The organizational amplifier
Individual planning fallacy is bad enough. Organizations amplify it:
Competitive pressure compresses estimates. When multiple teams or vendors compete for work, the most optimistic estimate often wins. Organizations select for optimism, then act surprised when optimism proves unfounded.
Layers add optimism. Each level of management may pressure for faster delivery. A realistic estimate from the team becomes optimistic by the time it reaches executives, with no additional information added, just pressure.
Coordination costs are invisible. Individual tasks might be estimated reasonably, but the time spent coordinating, waiting, handing off, and resolving conflicts rarely appears in the plan. These costs scale with organizational complexity.
Success theater hides reality. No one wants to report they’re behind. Status reports stay green until they suddenly go red. By the time leadership learns about problems, the planning fallacy has already done its damage.
Memory is selective. The successful projects get celebrated; the delayed projects get memory-holed. Organizational learning about estimation accuracy doesn’t accumulate because the organization doesn’t want to remember.
Approaches that actually help
The planning fallacy can’t be eliminated by trying harder to estimate accurately. It requires structural interventions:
Reference class forecasting. Don’t ask “how long will this project take?” Ask “how long did similar projects actually take?” Use the base rate from comparable efforts as your starting point, then adjust for specifics. This forces outside-view thinking rather than inside-view optimism.
Pre-mortems. Before starting, imagine the project has failed and ask: “What went wrong?” This exercise surfaces risks that optimistic planning overlooks. It’s easier to imagine specific failures after you’ve already (hypothetically) failed than to imagine them when you’re envisioning success.
Buffer explicitly. Add contingency that isn’t allocated to specific tasks. Not padding hidden in estimates, but explicit buffer for the unknown unknowns. When someone asks why there’s buffer, the answer is: “Because our historical estimates have been optimistic by X%, and we’re building that learning into this plan.”
Track and learn. Compare estimates to actuals, systematically, over time. Build organizational memory about estimation accuracy. Which types of projects run long? Which estimators are consistently optimistic or pessimistic? Data beats intuition.
Separate planning from commitment. Estimate what the work will actually take, then decide whether to commit. Conflating “what will this take?” with “what are we willing to invest?” produces estimates that are really negotiations.
When optimism serves a purpose
The planning fallacy isn’t entirely dysfunctional. Some optimism is necessary:
Bold projects wouldn’t start. If everyone accurately understood how hard a project would be, some worthy efforts would never begin. Founders, pioneers, and innovators benefit from optimism that lets them attempt things others wouldn’t.
Motivation matters. Teams that believe they can hit a target work harder than teams that believe they can’t. Pessimistic estimates, even if accurate, can become self-fulfilling prophecies.
Not everything is plannable. For genuinely novel work, historical base rates don’t exist. Some degree of optimistic estimation is unavoidable when you’re doing something that hasn’t been done.
The goal isn’t to eliminate optimism but to balance it with realism. Optimism should fuel commitment to worthy goals; realism should shape how resources are allocated and expectations are managed.
The organization that expects to launch in six months but resources for twelve is better positioned than the organization that expects six and resources for six. The former has room to adapt; the latter faces crisis when reality arrives.
Living with the fallacy
Perfect estimation isn’t possible. The planning fallacy reflects deep features of human cognition that can’t be engineered away. But planning can improve:
Expect overruns; plan for them. Build the assumption of delay into how you sequence work, allocate resources, and communicate with stakeholders.
Create checkpoints. Don’t wait until the end to discover the plan was wrong. Frequent milestones reveal divergence early enough to respond.
Embrace range estimates. A single number (“it will take six months”) is almost certainly wrong. A range (“it will take six to nine months, with nine being more likely based on similar projects”) is more honest and more useful.
Reward accuracy, not optimism. If the organizational culture celebrates aggressive timelines and punishes realistic ones, optimistic estimates will continue. Change the incentives; change the estimates.
The planning fallacy humbles us. We are not as good at predicting the future as we think we are, especially our own future. The best response isn’t to pretend otherwise. It’s to build planning processes that account for our predictable unpredictability.
Strategic Advisory helps organizations develop planning processes grounded in reality, using reference classes, explicit buffers, and learning systems to counteract optimistic bias.
Intelligent Operations builds the project tracking and milestone systems that surface divergence from plan early, when there’s still time to respond.