This is untimed educational practice for one applied skill. It is not a clinical, diagnostic, employment, or professionally recognized assessment, it does not produce a score or credential, and for the judgement and safety families it is not a substitute for your employer's own policies, training, or legal duties.
Work that unblocks other people has a multiplier the calendar does not show: two minutes of yours buys back hours of theirs, and the board paper is still comfortably deliverable afterwards. Starting with the board paper is the intuitive answer because seniority feels like priority, but it is one person's output blocking nobody, while the approval blocks two.
Delegating with a clear outcome and a checkpoint is how a team's capacity stops being capped by one person's working hours, and the early slowness is the cost of buying that. Keeping the critical pieces is the reasonable-sounding compromise, and it is the version that never ends: the critical work is exactly the work on which people learn to be trusted.
Delegating an outcome rather than a method is what makes the person responsible for the result instead of for following you. Requiring your approach is defensible when consistency is genuinely a requirement, but here it is not stated as one, and once you own the method you own every decision inside it, which is the opposite of delegating.
A straight answer plus whatever is genuinely available keeps the conversation honest and gives them something real this year. Promising next year is the kindest-sounding option in the room and the most damaging: you do not control that budget, and an undelivered promise is what turns a motivated employee into a leaving one.
Slack is exactly when capability is cheap to build, and a second capable person is what protects the schedule the next time the first one is unavailable. Assigning for speed is the efficient answer for this month and the reason teams end up with a single point of failure for the task nobody else has ever done.
A three-week change in a previously reliable person usually has a cause (a blocker, a workload change, something outside work), and a conversation is the only way to find out which. Opening with the formal plan is the option that looks most procedurally correct, but starting formal before you know the cause makes the honest explanation much less likely to be offered.
Specific behaviour, specific effect, specific request is what someone can actually act on, and doing it promptly and in person allows them to respond. The compliment sandwich is the most widely taught alternative and the least reliable: people routinely leave the room remembering the praise and unsure that anything needed to change.
Outcome and contribution are two different records, and keeping them separate means the business still knows what happened while the individual is judged on what was theirs to influence. Awarding the counterfactual rating is the sympathetic option, but it puts a result in the record that did not occur, and next year nobody can tell which ratings describe reality.
Written criteria applied to evidence is what stops the assessment tracking who talks most, and visibility is then coached rather than silently priced in. Rating visibility higher is the plausible objection, and communication genuinely can be real value, but if it matters it belongs in the criteria for everyone in advance, and if it is not in the criteria then rewarding it is a bias rather than a measure.
How someone gets results is part of their performance, and two independent reports make it a pattern rather than a personality clash. Tolerating it is the tempting trade because their output is genuinely valuable, but the cost lands on other people's output and on who stays, and none of that appears in the high producer's numbers.
The person carrying the cost should hear it first, in a setting where they can ask questions and react without an audience. Announcing it to everyone at once is the plausible alternative and has a real fairness logic behind it, but it makes the affected person receive bad news in public, and that is the part they will remember about how you manage.
Disclosure is the act that protects everyone: it puts the potential partiality on the record and lets the process, not you, decide whether you can still be involved. Standing aside unilaterally is the scrupulous-sounding option, and a complaint left with nobody assigned can sit unowned for weeks. It is not usually the manager's call to make alone.
Naming it, fixing it, and changing the mechanism deals with the individual harm and the next one at the same time. Silently correcting the outcome is the option that avoids an uncomfortable conversation, but the person still believes your original judgement of them stands, and that belief is most of the damage.
Setting the limit before they speak means their decision to tell you is an informed one, and it keeps you able to act if what follows is something you cannot sit on. Agreeing outright is the answer that feels most supportive in the moment, and it is the one that ends with either a broken promise or an obligation you have disabled yourself from meeting.
The team needs to understand the reasoning, know who is accountable for delivery, and have a route for their objections to travel. Distancing yourself is the honest-feeling option, but it leaves a decision that nobody in the room owns, which converts specific concerns into general grumbling with nowhere to go.
Owning the outcome and examining the system produces changes that survive the next staff change, and it keeps people willing to report the next problem early. Adding an approval step is the fastest-looking control and often the wrong one: introduced before the cause is understood, it adds delay to every future release without addressing the failure that actually occurred.
A stated deadline, one clear hearing for each case, and a recorded rationale lets people disagree and still commit, because they can see the argument that carried. A vote is the fairest-looking mechanism and the weakest here: it settles a technical question by headcount and leaves the losing half with an outcome but no reasons.
Estimate plus assumptions plus options turns a demand into a decision the stakeholder can make with open eyes, and it is the only version where the date they choose means something. Committing and pushing is the answer that reads as can-do, and it converts a scheduling problem you could solve today into a credibility problem you cannot solve later.
Money already spent is gone whichever way the decision goes, so the only question worth asking is what the remaining spend buys. 'Stopping now wastes everything already invested' is the most natural thing to say in the room and the clearest form of the sunk-cost error: it treats past spending as a reason to continue rather than as information about the original estimate.
Being explicit about what is fixed and what is open lets people influence the part that can move, and showing what their input changed is what makes the next consultation credible. Introducing it gradually and unannounced is the option that avoids resistance today; when the team works out what happened, and they will, every future change is met with suspicion first.