What it means
The bias comes from how we build forecasts in the first place. We picture a specific plan running to plan, and the things that derail projects, a key person leaving, a supplier failing, a permit taking three months instead of three weeks, are each individually unlikely and so get left out.
Add all those small risks together and the chance that none of them bites is actually quite low. The result is an estimate that is not merely uncertain but systematically skewed in one direction, which is what makes it a bias rather than noise.
The business cost is real and measurable. Capital projects overrun, product launches slip, and cash forecasts that looked comfortable turn tight, which forces emergency borrowing at exactly the moment when terms are worst.
The standard correction is reference class forecasting: instead of asking how long this project will take, look at how long genuinely similar projects actually took and apply that track record as an uplift. Some organisations formalise this by adding a fixed contingency percentage derived from their own history of overruns.
Optimism bias is not the same as deliberately under-costing a project to get it approved, though the two often travel together. The practical test is simple: if your last five projects all overran and your next estimate assumes this one will not, the bias is still sitting in the number.
The cure is structural rather than personal, because willpower does not fix a bias you cannot see. Organisations that estimate well tend to keep records of what past projects actually cost and require every new proposal to be presented against that history.
In practice
Real-world examples.
Example
A marketing director forecasts that a new campaign will generate 500 qualified leads a month from launch. The finance team applies the company's own history, where new channels reached about 60% of forecast in the first quarter, and plans hiring against 300 leads instead.
Example
A construction client is told a refurbishment will take fourteen weeks. Knowing that similar fit-outs in the same building took eighteen to twenty-two weeks, the client negotiates a two-month lease overlap rather than committing to move out on week fifteen, and budgets $40,000 for the extra rent.
Example
A founder models a fundraising round closing in eight weeks and plans payroll accordingly. An adviser points out that comparable rounds typically take four to six months from first meeting to money in the bank, so the founder trims discretionary spending early instead of running out of cash halfway through the process.
Think of it
“Optimism bias is expecting the best-overestimating good outcomes, underestimating bad ones.
Formula
Calculation
Adjusted forecast = base estimate x (1 + historical overrun rate)
An engineering team estimates that a new production line will cost $800,000. Reviewing the last eight comparable projects, the finance director finds that actual costs came in an average of 40% above the original estimate.
Applying that track record, the adjusted forecast is $800,000 x 1.40 = $1,120,000, which means the uplift for optimism bias is $320,000. The board approves a budget of $1,120,000 up front rather than discovering the shortfall halfway through the build, when stopping is no longer an option.Case study
Seen in the real world.
Brightwater Labs is a fictional diagnostics company invented to illustrate how optimism bias compounds. Its team forecast that a new testing platform would cost $2,000,000 and be selling within twelve months, with revenue of $1,500,000 in the first year of sales.
The illustrative reality was that regulatory review took seven months rather than three, one supplier missed a deadline, and the launch slipped by nine months. Costs reached $2,900,000, a 45% overrun, and first-year revenue was $600,000 because the sales team had been hired against a launch date that never happened.
Reviewing the wreckage, the fictional finance director introduced a rule: every project estimate above $500,000 must be presented alongside the actual outcome of the three most similar past projects. Estimates did not become pessimistic, but they stopped assuming that everything would go right at once.
Watch out
Common mistakes.
- Assuming optimism bias only affects inexperienced teams. Experts are often more confident and therefore more exposed, because their expertise makes the base plan feel more certain.
- Correcting for it by simply adding a round 10% to everything. If the organisation's real overrun history is 40%, a token contingency just makes the same error look considered.
- Confusing optimism bias with a positive attitude that should be encouraged. Optimism is useful for motivation and damaging in a forecast, and the two can be kept separate.
Questions
People also ask.
How do I spot it in my own numbers?
Compare your last several forecasts with what actually happened; a consistent one-directional gap is the signature of bias rather than bad luck.
Does a bigger contingency fix the problem?
Only if the contingency is sized from real historical data and is protected from being spent on scope creep the moment it is available.
Is pessimism the answer?
No, the goal is calibration, and deliberately pessimistic estimates cause their own damage by killing projects that would have paid off.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%