The sponsor asks how long the work will take. You have a reasonable idea, but the answer depends on inputs nobody has confirmed. Give a project estimate as a range by describing the work, stating the conditions behind the lower and upper ends and explaining what evidence will narrow it.
A range should expose uncertainty, not disguise a commitment. APM's estimating guidance addresses uncertainty; the record below is a practical editorial tool rather than a statistical confidence model.
Photo by Adrian Regeci on Unsplash.
Define the result being estimated
State what completed means and which preparation, checking and correction are included. An estimate to produce a draft is different from an estimate to obtain an accepted deliverable.
Separate active effort from elapsed duration. Record who supplies inputs, who reviews the result and when those people are available. If the scope is unclear, give the next discovery action and its estimate instead of attaching a precise duration to an undefined whole.
Explain both ends of the range
Try this short estimate record:
Deliverable, boundaries and acceptance condition.
Expected work and main sources of uncertainty.
Lower duration and the conditions needed to achieve it.
Upper planning duration and the difficulties it allows for.
Evidence used, relevant past comparisons and limitations.
Inputs or decisions that would change the range.
Owner, review date and intended use of the estimate.
Do not describe the upper end as a worst case unless you have genuinely considered the relevant extremes. A difficult-but-plausible scenario is usually a more honest label for a practical planning discussion.
An illustrative estimate conversation
A team estimates that preparing an internal service guide will take three to six working days once the approved process is available. Three assumes complete source material and one consolidated review. Six allows for resolving specific missing examples and another review cycle.
The process itself has not yet been approved. That waiting time sits outside the stated range and must be shown separately in the delivery forecast. Otherwise the sponsor may hear “six days from today” when the team means “six days after a condition is met”.
The figures are hypothetical. They illustrate how explaining conditions can be more useful than averaging two guesses into one confident number.
Update the estimate when the evidence changes
After the first review, record what remains and reassess the remaining work. Do not preserve an obsolete range to avoid an uncomfortable conversation. Equally, do not reset the history so earlier assumptions disappear.
A changed estimate should identify its cause: different scope, more work revealed, an input failure or an original estimating error. Those explanations lead to different responses. Keep the forecast current while following the proper approval route for any changed baseline or commitment.
Agree what decision the estimate supports
An early range may support a discovery budget or a provisional slot. A firm client promise may need more evidence and an authorised contingency decision. Avoid presenting the same uncertain estimate differently simply because one audience wants certainty.
The small-team project plan provides the place to connect estimate assumptions to dependencies. For the connected task, read How to Keep a Project Assumptions Log That Changes Decisions. If estimates keep becoming promises before the conditions are understood, discuss your project requirements.
