Start with binary completion
For simple projects, count completed actionable items divided by total actionable items. This is transparent and easy to audit.
Do not count vague notes or reference material as work items. The denominator should represent things that can actually become complete.
Use weights when tasks differ greatly
Ten tiny tasks and one critical deliverable should not necessarily contribute equally. Assign simple weights only when the difference is meaningful. Excessive precision creates maintenance work without improving the decision.
A small scale such as 1, 2, and 3 is often enough.
Track milestone state separately
Milestones describe whether the project has crossed important boundaries. A project may be 80% complete by task count but still unable to submit because the final approval milestone is blocked.
Show both numerical progress and the next unresolved milestone when decisions depend on it.
Do not reward starting
A task that is 90% done but cannot yet be delivered is still incomplete for many planning decisions. If partial progress matters, define explicit substeps instead of repeatedly guessing a percentage.
Observable states make collaboration and review much easier.
Use progress to choose the next action
Progress is useful only if it changes what you do. Highlight blocked work, due-soon milestones, and the next available action rather than celebrating a number by itself.
A trustworthy 55% is more useful than an optimistic 90%.
Keep fixed dates on the calendar, executable actions as tasks, and multi-step outcomes inside projects. Preserve source evidence whenever a date came from a document.
Choose a metric that matches the decision
If the question is whether the project can ship, milestone state may matter more than task count. If the question is how much routine work remains, completed weighted tasks may be more useful. There is no single percentage that is correct for every management decision.
Display the metric together with the next blocked or available action so the number stays connected to execution.
Audit the denominator
Progress becomes misleading when the task list changes but the denominator does not. When new required work is discovered, add it to the project rather than preserving an attractive percentage. A temporary drop in progress can be a more truthful description of the project.
Archive optional ideas separately so they do not permanently inflate the amount of required work.