COST-TO-COMPLETE IS ALWAYS WRONG.
Cost-to-complete estimates are structurally biased optimistic because the person estimating remaining cost is usually the same person whose performance the number reflects. The fix isn't a better guess — it's a WIP process that forces the estimate to be checked against production data every month.
Every job has a cost-to-complete number, and almost every one of them is wrong in the same direction: too low. That's not bad luck or bad estimators. It's structural — the field wants the job to look on track, the PM doesn't want to deliver bad news, and the number that gets reported is the number that keeps everyone comfortable, not the number that's actually true. Left unchecked, this is exactly how profitable-looking jobs turn into loss jobs at closeout, and how WIP schedules quietly overstate earned revenue for months before anyone catches it.
THE PERSON GUESSING HAS A REASON TO GUESS LOW.
Cost-to-Complete Is Self-Reported
The PM who reports cost-to-complete is usually the same person whose job performance that number reflects. Reporting a higher cost-to-complete means admitting the job is running worse than planned — so the number quietly drifts toward "fine" instead of "accurate."
Optimism Compounds Every Month
A cost-to-complete estimate that's slightly low in month three gets used as the baseline for month four's estimate. Each month's optimism stacks on the last, so the gap between "reported remaining cost" and "actual remaining cost" grows quietly until closeout forces it into the open.
WIP Reporting Inherits the Error
Percentage-of-completion accounting uses cost-to-complete to calculate earned revenue. An optimistic cost-to-complete number doesn't just mislead the PM — it overstates earned revenue on the company's own financials, which is how a WIP schedule can look healthy right up until a job closes out underwater.
CHECK THE ESTIMATE AGAINST PRODUCTION.
THE WIP SCHEDULE IS ONLY AS HONEST AS THE INPUTS.
A WIP schedule doesn't lie on its own — it just faithfully reports whatever cost-to-complete numbers get fed into it. If those numbers are structurally optimistic, the WIP schedule will be too, and the first place that shows up is usually a job that "looked fine" right up until it closed out losing money.
Fixing this isn't about better estimators. It's about a monthly process that checks the estimate against real production data instead of trusting a number that has every incentive to be too low.
