Start with a project you can define.

Define the outcome, then the deliverables

An outcome describes the change a project should create. A deliverable describes something you can hand over. “Improve onboarding” is an outcome; an approved onboarding checklist and a revised welcome sequence are deliverables. Keeping these separate prevents a broad ambition from quietly becoming an unlimited commitment.

For each deliverable, record its format, intended audience, responsible reviewer, and acceptance criteria. Instead of “complete the design,” write “supply the agreed page layouts at the agreed screen sizes, with one consolidated review round.” Specificity helps both parties spot omissions before those omissions become disputes.

Break the work into estimable pieces

A task should be small enough that somebody can explain how they will do it and what evidence will show it is complete. Large labels such as “development” hide uncertainty. Separate setup, implementation, checking, revisions, and handover so their effort remains visible.

Excessive detail has a cost too. A board with hundreds of tiny administrative actions becomes difficult to maintain. Split a task when it has a different owner, a distinct approval point, a meaningful dependency, or unusually uncertain effort. Combine routine actions when tracking them separately would not change a decision.

• Include preparation and access requests before production.
• Include internal review before client approval.
• Include documentation, delivery, and final administration.
• Keep client waiting time separate from working hours.

Choose an estimating approach

Bottom-up estimating adds the expected effort for individual tasks. It is useful when the scope is reasonably clear and you need to explain the budget. Its weakness is omission: a detailed calculation still fails if a necessary task never makes the list.

Top-down estimating starts with a total budget or delivery window and allocates it across the work. It is faster for early feasibility checks, but can force unrealistic expectations onto individual tasks. Use it as a constraint, not evidence that the work actually fits. Comparing both approaches exposes gaps worth discussing.

Make assumptions visible

An assumption is a condition your estimate relies on, not a fact you have confirmed. Examples include timely access to systems, usable source files, a single decision-maker, and feedback arriving within an agreed review window. Record who will confirm each assumption and when.

When uncertainty is substantial, describe a range outside the calculator and enter a clearly labeled working estimate. A discovery task can reduce uncertainty before you commit to the rest of the work. Avoid adding an arbitrary contingency simply to conceal a scope question that needs an answer.

Control changes without making collaboration difficult

Define what happens when somebody requests another deliverable, an extra review round, or a shorter deadline. A proportionate change process records the request, its effect on cost and schedule, and the approval needed before work begins. It need not be bureaucratic to be useful.

Keep the original estimate as a baseline by exporting a summary before changing the live plan. PlanForge does not maintain a version history. Date your exported files and keep approvals in your usual document system so you can explain how the project evolved.

Put the brief into PlanForge

Create a descriptive project name, set the overall date window, and add the tasks that deliver your scope. Assign estimated hours and delivery cost rates, then add non-labor costs separately. The timeline can reveal crowded periods, but it does not automatically identify dependencies or overallocated people.

Review the plan with whoever will perform the work. Ask which task feels least certain, what must arrive from the client, and what could stop progress. The best first plan is not the most elaborate one. It is the one whose assumptions are clear enough to challenge and improve.