When Should You Escalate a Problem at Work?
There is a particular kind of stalling that happens right before someone raises an issue with a person above them. The message gets drafted and deleted. The meeting request gets cancelled twice. The problem sits there, half handled, while the person closest to it runs through every reason it might be better to keep managing it quietly a little longer.
Most of that hesitation comes from a single belief: that asking someone senior to step in is an admission that you could not do your job. It rarely gets said out loud in those terms, but it shapes the decision anyway. So people wait. They keep trying to resolve things within their own authority even after the situation has moved past what that authority can actually cover, or they swing the other way and flag anything faintly uncertain, hoping that volume will substitute for judgement. Neither pattern is really about the problem in front of them. Both are about managing how competent they appear.
That framing gets the whole question backwards. Escalation was never really a rescue request. It is a decision about where risk should sit, and recognising that changes what the hesitation is actually costing you.
The pattern that gets remembered
Two programme managers, working in similar roles at the same organisation, developed very different escalation habits over their first year. One raised almost everything that felt uncertain, framing each conversation around what she could not resolve on her own. The other escalated three times in twelve months, and each time arrived with the situation already mapped: what the problem was, what it would cost to let it continue, what the options were, and which one he recommended.
The frequent escalator was eventually asked directly whether her programmes needed additional resourcing. They did not. The projects were on track. What she actually needed, her director told her, was to start making the calls that sat within her own authority rather than routing them upward as a matter of habit. Within a few months her role had been restructured into something narrower, with a senior manager positioned above her to absorb the decisions she kept forwarding.
The other manager was promoted to oversee every strategic cross-functional initiative in the department. His director was specific about why: the escalations were rare enough that each one represented a genuine threshold, early enough that the decision could still change the outcome rather than just contain the damage, and clear enough that he could decide in a single conversation instead of chasing follow-up questions. The pattern itself, more than any individual save, had built the trust.
Neither of these people was being judged on how many problems they solved alone. They were being judged on whether they could tell the difference between a problem that belonged to them and one that had quietly become someone else's to own.
Where the threshold actually sits
The useful question is rarely "can I still handle this." Most capable professionals can keep handling almost anything if they push hard enough. The better question is whether continuing to handle it alone is now more expensive, in risk terms, than the ten minutes it would take to interrupt someone with the authority to accept that risk.
A finance manager once received an urgent request to move a substantial sum out of operational reserves and into a client-facing initiative launching in two weeks. Three department heads had already agreed to the transfer among themselves, treating it as a routine internal move well within their combined authority. On paper, it was. What none of them knew was that those particular reserves were tied to a staffing commitment made to a strategic client eighteen months earlier, one that carried an uncapped financial penalty if service levels slipped during the coming peak season.
She did the arithmetic. The exposure ran into penalties that could exceed what the new initiative itself was projected to earn. This was not a matter of preference or of being cautious for its own sake. Proceeding without flagging it created a specific, calculable loss that none of the people approving the transfer had the information to weigh. She wrote a two-page brief: the commitment, the exposure, the timeline before the shortfall would bite, and an alternative funding source that avoided the problem entirely. The reallocation was rejected in a fifteen-minute call. The client never knew how close the organisation had come to breaching the agreement.
That is the shape of a genuine threshold. Something is about to become difficult or expensive to undo, and the person best placed to see that has information the decision-makers do not. It is not about difficulty of execution. Plenty of hard problems belong entirely within someone's own remit. It is about whether the cost of being wrong now sits above what that person's role is actually meant to absorb.
The same logic held for a compliance officer who noticed that a data-processing pilot, running smoothly in one region, would breach data protection requirements the moment it scaled globally. Business teams wanted to roll it out immediately, and nothing about the pilot itself looked like a problem. The exposure only became visible once someone traced what happened at scale. She escalated before deployment rather than after, while the workflow could still be redesigned rather than unpicked. When a competitor was later fined heavily for a comparable failure, her early flag was the reason she was pulled into every major operational redesign that followed.
In both cases, the pattern is the same. The threshold is not a fixed number of pounds or a fixed number of weeks. It moves with the role, the organisation and how much risk that particular decision is allowed to carry before someone with broader authority needs to see it. Anyone looking for a universal trigger, a rule that applies identically in every company, is looking for something that does not exist. What does transfer between organisations is the underlying question: is this still reversible, and does the person handling it right now hold enough authority to own what happens if it is not.
What makes an escalation land well
The difference between an escalation that builds trust and one that erodes it usually comes down to what arrives alongside the problem. A vague warning that something might be wrong forces the senior person to do the analytical work themselves, which is precisely the burden escalation is supposed to lift from them. A well-framed one does that work in advance: what the issue actually is, what it will cost if nothing changes, what the realistic options are, and what the person raising it recommends.
That structure matters more than tone. It signals that the person is not asking to be told what to think. They have already thought it through, and they have identified the one part of the decision that genuinely requires authority they do not personally hold, usually the acceptance of a risk, a cost, or a trade-off between two legitimate priorities. Everything else in the situation, they are still handling themselves.
This is also where the earlier instinct to escalate everything or nothing falls apart. A useful working rule is to ask, before raising anything, whether the decision is still reversible within your own authority, and whether delay meaningfully increases the cost of being wrong. If the answer to both is yes, it is usually still yours to manage. If either answer flips, the conversation is worth having, and it is worth having with the analysis already done rather than the problem simply handed over half formed.
Responsibility, in other words, does not disappear the moment authority moves upward. The person who escalated well is still the one who framed the issue accurately, still the one whose judgement about what mattered turned out to be correct, and still the one whose name gets attached to the outcome when it goes right. What changes is who carries the decision itself, not who understood the situation first.
Escalation, looked at this way, stops being a measure of what someone could not do alone. It becomes one of the clearest signals an organisation has of whether a person's sense of risk can be trusted, which is precisely the kind of judgement that determines what gets handed to them next.
David Taylor
Interested in going further?
The AI-Ready Career develops this same judgement, over when to hold a decision and when to move it upward, into a wider framework for building a role that becomes harder to replace rather than easier to route around.
Ad · Amazon affiliate link.
