Estimate Like a Professional: A Field Checklist to Stop Guessing
Most schedule and cost overruns do not start in execution. They start the moment someone writes down a number nobody can defend. "It'll take about two weeks" feels harmless until two weeks becomes six, the budget is gone, and no one can reconstruct how the figure was reached. Good estimating is not fortune-telling. It is a repeatable method for producing a number, recording the assumptions behind it, and stating how confident you are.
After two years of pandemic recovery, hybrid teams, and supply disruption that is still fresh in everyone's memory, the old habit of estimating from a quiet, predictable baseline simply does not hold. Lead times wander, people are split across time zones, and a single delayed shipment can ripple through a plan. The checklist below will not make uncertainty disappear, but it will make your numbers honest, traceable, and easier to defend when someone asks how you got them.
Pick a method that fits what you actually know
There is no single right technique. The trick is to match the method to the quality of information you have, and to be open about which one you used.
Analogous estimating. Base the number on a similar past project. It is fast and useful early on, but only as good as the comparison — confirm the old work was genuinely alike before you trust it.
Parametric estimating. Use a measured unit rate — cost per square metre, hours per drawing, days per inspection — and multiply by scope. It scales well when you have reliable rates from real data.
Bottom-up estimating. Break the work into small pieces, estimate each, and roll them up. It is the most accurate and the most time-consuming, so reserve it for the parts that carry the most risk or money.
Three-point estimating. For each item, capture an optimistic, most-likely, and pessimistic value, then weight them (a common formula is optimistic plus four times most-likely plus pessimistic, divided by six). This forces uncertainty into the open instead of hiding it in a single tidy figure.
The field checklist you can use this week
Before you commit any estimate to a plan or a client, walk it through these steps. None of them take long, and together they catch the errors that quietly sink projects.
State the scope you are estimating in one sentence — if you cannot, the estimate is premature.
Write down every assumption: lead times, who is available, what is excluded, and what must be true for the number to hold.
Estimate the work, not the worker — base hours on the task, then adjust for the team you actually have.
Use ranges, not single points, for anything genuinely uncertain, and say which method produced each number.
Add a contingency that reflects real risk, and label it as contingency — never bury it inside the line items.
Have a second person who knows the work review the figures before they go out.
Record the date and the basis of estimate, so when reality differs you can learn instead of argue.
Treat every estimate as a learning loop
The single fastest way to improve is to compare what you estimated against what actually happened, and to do it honestly. Keep the original basis of estimate, log the actuals when the work is done, and look at the gap without blame. Over a few projects you will discover where your team consistently runs optimistic — testing, approvals, and integration are the usual culprits — and you can correct for it. An estimate that is documented can be challenged, improved, and reused. An estimate that lives only in someone's head can only be argued about.
If your team is making major delivery commitments and wants estimates that hold up under scrutiny, XNM's program & project delivery advisory can help you build a repeatable estimating discipline.