Skip to main content
Back to blog

Compare Estimated and Recorded Project Hours with AI

Review project hours in Clanker Workspace against an explicit estimate. Show phase-level differences, incomplete records, and scope questions before making a delivery decision.

When a project takes more effort than expected, the first useful step is a clear comparison. Which phase differs from the estimate? Are all hours recorded? Did the scope change? Those questions deserve evidence before you revise the plan.

Clanker Workspace can help prepare that review from an approved estimate, a dated hours report, and scope notes. The result should show what the records establish and what the project owner still needs to clarify.

Choose a comparable baseline

Put the estimate and hours report in the same Project. Identify the estimate version, phase definitions, reporting cutoff, and whether recorded hours are cumulative or only for one week.

Do not compare a whole-project estimate with a weekly export as though both cover the same work. Keep hours distinct from costs and billing. If the business needs a cost comparison later, supply the relevant rate definitions and scope separately.

Add approved scope changes and requests still awaiting a decision. An email requesting another review round can help explain a question, but does not prove that the extra round was accepted.

Try a phase-level example

This fictional table compares phase estimates with cumulative recorded hours through 13 October 2026. A blank recorded value means unreported, not zero:

phase,estimated_hours,recorded_hours
Discovery,8,6
Draft,16,20
Review,8,

The fictional notes say that an additional review round was requested, but approval is missing. The Draft phase is still in progress.

Ask:

Prepare project-hours-review.md from the estimate, recorded-hours table, and scope notes. Use difference = recorded hours minus estimated hours for phases with both values. Keep unreported hours unknown. Show the comparison's coverage and cutoff. Separate requested scope changes from approved ones. Do not treat the current records as a completion forecast or decide what to bill.

Open the resulting file and inspect the calculations alongside the source. Ask for an input list if the Agent uses a different estimate version.

Check the difference and its limits

The sample supports these findings:

Phase Supported finding
Discovery 6 recorded versus 8 estimated: difference −2 hours.
Draft 20 recorded versus 16 estimated: difference +4 hours; work is still in progress.
Review Recorded hours are missing, so its difference is unknown.

Across the two phases with reported values, 26 hours are recorded against 24 estimated: a net difference of +2. The full estimate is 32 hours, but total recorded effort remains incomplete because Review is unreported.

This distinction prevents the file from presenting 26 as a complete project total or claiming there are exactly six hours left. A phase below its estimate today is not automatically a finished phase, and current recorded effort is not a forecast of final effort.

Turn the review into questions for the owner

Ask which missing facts matter to the delivery decision. In the example, the owner should confirm Review's recorded hours, remaining Draft work, and whether the additional round is approved.

Keep proposed decisions separate from the comparison itself. The owner can resolve scope, update the estimate when appropriate, and record the reason for the new baseline. Save the earlier review when it explains how the plan changed.

For the next reporting period, use a current export with the same definitions and cutoff rule. A weekly report can include this comparison once the input process is reliable. The ongoing Project walkthrough helps preserve sources and reviewed outputs between visits.

Sources

Put it to work in Workspace

Give your assistant one useful job

Keep the reporting period, source evidence, blockers, and next decisions clear.

Try WorkspacePrepare a delivery review