Published October 12, 2026
Published October 12, 2026
Stop treating every task as equal
A list of twenty tasks may represent an easy morning or an impossible week. Capacity depends on effort, deadline, switching cost, uncertainty, and the amount of judgment required. Build the signal around those conditions rather than raw count.
Use lightweight size bands based on familiar examples. The goal is consistent planning, not precise time accounting. Recalibrate when the same class of work repeatedly differs from its estimate.
Separate ready, blocked, and recurring work
Ready work can consume capacity now. Blocked work consumes follow-up attention but may not be executable. Recurring responsibilities reserve capacity even when their tasks have not appeared in the queue.
Showing these lanes prevents an owner from filling apparent space that already belongs to a weekly close, client update, or scheduling cycle.
Add deadline pressure
Mark fixed external deadlines separately from preferred internal dates. Then show the next few consequences, not every due date. A capacity conversation improves when it names what will move or degrade if new work is accepted.
Identify dependencies and decision deadlines. Work that takes one hour after approval may still be at risk if the approval arrives too late.
Use a three-state signal
Available means planned work fits with reasonable contingency. Near limit means new work requires an explicit tradeoff. Over limit means current commitments already exceed the safe plan. Define these states with examples from the actual role.
Allow the assistant to raise the signal without proving exhaustion. Early visibility is the control. If people wait until missed deadlines make overload undeniable, the signal arrives too late.
Pair the signal with options
When capacity is tight, present choices: defer a lower-priority outcome, narrow scope, move a deadline, transfer work, add temporary support, or remove a blocker. “I’m busy” gives the owner little basis for a decision.
The owner should select the tradeoff and record it. Quietly asking the assistant to “do their best” leaves all commitments apparently intact.
Watch switching and hidden coordination
Rapid movement between inbox, scheduling, research, client service, and finance can reduce usable capacity even when each task looks small. Group compatible work and protect blocks for tasks requiring concentration.
Count coordination that the role genuinely requires: chasing inputs, preparing approvals, documenting exceptions, and confirming completion. These are part of delivery, not administrative waste around it.
Review forecast against reality
At the end of each cycle, compare planned and actual work. Look for recurring underestimates, emergency additions, blocked decisions, and unnecessary rework. Adjust the operating design instead of demanding better prediction from unstable inputs.
A credible capacity signal lets an assistant surface constraints without drama and gives the owner a concrete set of tradeoffs. It protects delivery quality by making overload a planning decision rather than a personal failure.
This guide was published on October 12, 2026. For adjacent workflows, explore the Delegation Assistant blog or review the deeper evidence in Research.