You keep adding work between plans. The customer escalation you personally promised to resolve. The report the board requested. The early work on next quarter's initiative. A manager who reports to you says yes, and the things you asked for arrive.
From your chair, it looks like the work fit.
Sometimes it did. Other times, a deadline quietly moved. A review got skipped. The manager spent another evening finishing work that no longer fit into the day. You saw the delivery, not what paid for it.
That's the failure worth looking for: the tradeoff moved downstairs. You chose the addition; someone else was left to choose what gave.
Before you read the next missed commitment as a performance problem, find out what changed when you added the work. And don't make the entire remedy an instruction for managers to push back harder. I'd put the burden on you to open the conversation and own the choices they aren't authorized to make.
Why a yes tells you too little
Raising a workload conflict can feel like admitting incapacity. A manager may assume you'll reject a request to delay something. They may also be remembering a real refusal. Ask which before treating the problem as a lack of candor.
When you keep receiving yes without hearing what has to move, you can mistake quiet coping for spare capacity. If the person who raises a conflict gets a lecture about commitment while the person who absorbs it gets praised, you've made silence the easier choice.
I'd be especially careful with that praise. You can thank someone for handling everything and still avoid the decision they need from you. Recognize the effort. Then ask whether delivering that way required a sacrifice you'd want repeated.
None of this makes a yes suspicious by itself. A manager with available capacity, a better method, or work that has just finished may have a perfectly sound yes. You don't need to force a subtraction to prove you're being disciplined.
Start with the work you added
Pick one manager carrying a recent addition and use your next one-on-one to trace what changed. If a commitment is already at risk, don't wait for that meeting. This is an operating check I'd try, not a validated test of overload.
Open with your own request. If you didn't discuss its effect on existing commitments, acknowledge that. Ask what happened to the other work when it arrived. Give the manager time to check rather than requiring an answer on the spot.
Check whether any deadline moved, an activity stopped, extra hours were needed, or another team was affected. Don't ask them to produce an overload story. Ask what changed.
A moved deadline or missed review is a reason to investigate, not proof that your request caused the problem. A dependency may have failed. The original estimate may have been wrong. The manager or team may need help developing a skill. Work through those possibilities rather than making overload the answer to every miss.
Test assumptions in both directions, too. Ask what the manager previously raised with you and what happened. If you've already rejected relief, don't reclassify that history as a confidence problem. Check your side: are you avoiding a clear tradeoff because you assume the manager can't handle the conversation?
Then look ahead. Compare the current workload with the outcomes you agreed mattered most. Ask the manager to explain the order they're using, what is waiting, and which deadline or quality standard would be at risk if the next substantial request arrived. Have them recommend what should change.
You can do this before anything slips. The useful question when assigning work is how it fits, not whether the manager can produce enough evidence of distress to earn a discussion.
Make the trade at the right level
Your manager should still own routine scheduling and choices within their authority. Don't turn this into a system where every small adjustment needs your approval.
But if making room requires changing a commitment outside their authority, they need more than permission to prioritize. That might be a customer promise you made or a priority you control. Make or authorize the choice, and involve any other leader whose commitment has to change.
Stop lower-value work, reduce scope, defer the new ask, or add help that can actually contribute in time. Ask the manager to compare the practical options, including what each one costs. Count the time the team will spend training someone before treating a new person as relief.
Consider a customer escalation that pulls the only available engineer off a release. You may decide the escalation comes first. Then settle what happens to the release: a different date, a smaller scope, or credible replacement coverage. If your promise created the conflict, don't leave the manager alone to explain the changed commitment to everyone else.
If you choose to proceed with a known delivery risk, write down the likely delay, reduced scope or other impact, and what the team can reasonably commit to. Best efforts cannot mean keeping an unchanged promise that you no longer expect the team to meet. Nor can accepting a cost waive safety or legal requirements.
The manager remains accountable for executing the revised commitment and raising new problems. You remain accountable for the priorities and promises you changed. Put the revised commitment in the delivery plan, not just your one-on-one notes. Don't approve a delay and then treat the old date as the performance standard.
Check what actually changed
A conversation doesn't create headcount. If the required work still exceeds what the team can deliver, you need more usable capacity or a smaller workload. A clearer list isn't a substitute for that choice.
Some seasons genuinely ask more of people. A short crunch needs an end date, limits and recovery time. Calling it temporary doesn't make an impossible load workable, and extending it should require a fresh decision rather than another appeal to resilience.
Your response helps determine whether raising the next conflict seems worthwhile. Don't penalize a good-faith warning. Explain the choice, back the manager with affected stakeholders, and follow through on the relief you agreed to provide. If earlier warnings were brushed aside, one welcoming conversation won't erase that history.
At the next one-on-one, or sooner if a commitment is at risk, check whether the postponed work actually moved, the promised help arrived, and the consequences stayed within what you agreed. If the evidence points to a different problem, change the response.
And if the task you supposedly stopped is still on that manager's list for Friday, the tradeoff is still downstairs.