How to include large irregular jobs in a workload forecast

01

List the jobs separately

Choose a planning period and identify large jobs that are confirmed, under discussion or possible. For each, record the expected start and delivery dates, stages, required skills and estimated hours. Show the source of the estimate and who can update it. Keep existing backlog separate from new work so an open job is not counted twice.

Split a job into stages where possible. A design phase may need different people and dates from delivery or approval. An unfinished scope should be marked uncertain; do not fill a gap with a precise number that nobody can explain.

02

Build scenarios that can be scheduled

Make a base plan with confirmed commitments and the routine workload. Add a plausible scenario for a possible large job, stating its assumed start date, effort and dependencies. If two jobs might overlap, include a scenario with both rather than assuming they arrive in different weeks.

For example, a team has 80 usable hours in a week and expects 50 hours of routine and existing work. A possible project would need 20 hours that week, leaving an estimated 10 hours. If another project also requires 25 hours, the combined scenario needs 95 hours, creating a 15-hour gap before unplanned work. The first project's possible status does not make its 20 hours a confirmed booking.

03

Decide what happens if work is confirmed

Compare each scenario by skill and date, not only the total hours. Identify which commitments can be moved, who can approve a revised deadline and when a customer must receive an answer. Set a decision date for reserving people or supplier time, and document the cost or consequence of doing so.

A probability-weighted average can help describe expected demand across many opportunities, but it cannot schedule a specific job. A job that arrives still requires its full resources. Keep the actual possible workload and timing visible beside any average.

04

Update as uncertainty changes

Review opportunities at a regular cadence and when scope or dates change. Move confirmed jobs into the commitment plan and remove lost opportunities from the scenario, keeping a record of why the forecast changed. Compare actual stage hours with estimates once work finishes.

If historical records include few large jobs, say so and ask the people who perform the work to review the range. Check whether subcontractors or equipment are available before treating them as extra capacity.

05

Keep a separate project register

Keep a project register alongside the routine forecast: expected start window, probability or status, hours by skill and the decision date. Note any dependencies. Compare a no-project scenario with the full project workload. A probability-weighted average may help estimate total expected demand, but it can hide a week in which the whole project arrives and needs real capacity.