A typical Fortune 500 company burns roughly 530,000 days of managers’ time every year on inefficient decision making, which McKinsey pegs at about $250 million in wages before you count the missed launches and lost market share. Executives told the same researchers they spend 37 to 40 percent of their hours making decisions, and 61 percent of them believe most of that time is wasted. Read those numbers as a product manager and one line jumps out: the expensive part is not the wrong decision. It is the decision nobody makes.
That is the case for escalation, and it is the part of the job most PMs treat as an admission of defeat. Two teams want opposite things. You have run the alignment meetings, written the brief, heard everyone out, and you are still stuck. The instinct is to keep grinding, to schedule one more working session, to believe that if you were just a little better at influence the deadlock would dissolve. Sometimes it does. Often you are simply postponing a decision that belongs one level up, and every day you postpone it, the sprint the team is waiting on drifts.
I spent 20-plus years in IT operations before doing fractional COO work, and I have watched capable managers sit on deadlocks for weeks because escalating felt like tattling. It is not. Escalation done well is one of the clearest signals that a PM understands the boundary of their own authority. The skill is not deciding whether to escalate. It is escalating in a way that gets a fast decision without torching the relationship with the person you are escalating past.
Escalation is a routing decision, not a verdict
The framing that keeps PMs stuck is treating escalation as a judgment about who was right. It is not. It is a routing decision. A choice sits above your pay grade to make unilaterally, so you move it to the person who owns that tradeoff.
Arne Kittler, a product leader who has written extensively on this, makes the point that pushing for clarity is a core PM job, and alignment includes creating clarity in dissent. When two people cannot resolve a conflict because any resolution compromises one side’s goals more than they can absorb, there is genuinely no answer at that level. The worst outcome, he argues, is not escalating and letting the decision rot.
I think of it the way I thought about a network incident in my telecom years in Saskatchewan. When a field tech hit something outside their authority to fix, escalating to the network operations center was not failure. It was the system working. The tech who tried to hero their way through a problem they were not authorized to solve was the one who caused the outage that made the newspaper. Product decisions are slower-motion versions of the same thing.
The reframe matters because it changes who you are in the room. If escalation is a verdict, you are the person who lost an argument and ran to a parent. If it is routing, you are the person who correctly identified that a decision needed an owner and made sure it got one. Same action. Completely different read.
Do the work that earns a fast yes
Escalation gets a bad name because most of it is done badly. Someone forwards a thread to a VP with “can you weigh in?” and now the VP has to reconstruct the entire disagreement, chase down both positions, and figure out what is even being asked. That is not escalation. That is delegating your homework upward, and it is the behavior that makes leaders dread the word.
A clean escalation does the opposite. Atlassian’s clean escalation play lays out a sequence I have used almost verbatim: acknowledge the disagreement openly, clarify the options with real pros and cons, understand the tradeoffs behind each side’s position, then take it up. The test for step three is whether each party can describe the other’s position accurately enough that the other side says “yes, that is what I mean.” If you cannot pass that test, you are not ready to escalate. You are still in the disagreement.
By the time you bring it up, the person deciding should get one page, not a thread. What is the decision. What are the two or three real options. What does each team recommend and why. What is the cost of continuing to wait. And critically: exactly what you need from them, whether that is a yes or no by Friday, a conversation with a peer, or a specific call on a tradeoff. Do not make the decider figure out what you want. Tell them. This is the same discipline behind a good stakeholder decision brief, aimed one level higher.
Atlassian suggests resolving most escalations inside three to five days. That number is only achievable if the prep work is done before the escalation lands, not after.
Never let the escalation be a surprise
Here is the move that separates PMs who escalate for years without making enemies from the ones who get quietly frozen out: they tell the other party first.
The line is simple. “It looks like we still see this differently, and I do not think we are going to close the gap ourselves. I want to take it up to our director to decide. Do you want to be part of that conversation?” You send that before you send anything up the chain. It costs you nothing and it changes everything, because the alternative is that your counterpart learns you escalated when their boss’s boss asks them about it in a hallway. That version reads as an ambush, and people remember ambushes.
Inviting them into the escalation also protects you. It removes any suggestion that you framed the disagreement to your advantage behind their back. Both positions get represented. The decider hears both sides. And the person you disagreed with watches you handle a conflict with them honestly, which is worth more to the working relationship than winning this particular call. I have escalated against people and had them thank me later, specifically because the process was transparent. That does not happen when you go around someone.
If you are early in the disagreement and have not exhausted the direct routes, this is also the moment to make sure you are not escalating something you could still resolve by naming the conflict directly. Escalate too early and you look like you cannot handle friction. Escalate too late and the business eats the delay. The judgment is in that gap.
What happens after the decision is the real test
The decision comes down. Half the time it does not go the way you argued. This is where most of the value either gets created or destroyed, and it has a name most PMs have heard and few practice: disagree and commit.
Jeff Bezos wrote it into Amazon’s leadership principles and described it plainly in his 2016 shareholder letter. He greenlit an Amazon Studios show he personally doubted, telling the team his view, then writing back “I disagree and commit and hope it becomes the most watched thing we’ve ever made.” The point is not that he caved. It is that once the decision was made by the people who owned it, he committed for real, not with the quiet reservation of someone waiting to be proven right.
When a decision goes against you after an escalation, the team is watching how you carry it. If you deliver it with an implied “I told you so” or a shrug of “leadership decided, not me,” you have taught everyone that your commitment is conditional on getting your way. That poisons the next ten decisions. If you deliver it as a genuine call you now own, you have kept the thing escalation is supposed to protect, which is the team’s ability to move together after a hard choice. This is closely related to delivering setbacks without losing stakeholder trust: the discipline is owning an outcome you did not personally choose.
Disagree and commit only works if the disagreement was real and heard. You cannot skip the arguing and jump to the committing. But once a decision has an owner and that owner has ruled, relitigating it in Slack for the next two weeks is just the slow, expensive version of the deadlock you escalated to end.
The quiet cost of not escalating
The McKinsey research found that organizations able to make decisions both fast and well deliver financial returns roughly double their slower peers. Speed and quality are not a tradeoff at the good ones. The blocker is almost never that the right answer is unknowable. It is that the decision has no clear owner and everyone is too polite, or too worried about looking like they failed, to route it to one.
That is a PM problem, and it is a communication problem more than an analytical one. Knowing when a decision has left your authority, packaging it so a busy person can decide in one read, telling the other side before you go up, and committing hard to whatever comes back: none of that is on the standard PM competency list, and all of it is what actually keeps a product moving when two good teams want two different things.
The next time you catch yourself scheduling a fourth alignment meeting on the same stuck decision, ask a blunter question. Is this mine to decide? If it is, decide it. If it is not, the most useful thing you can do this week is get it cleanly, honestly, and quickly into the hands of the person whose call it actually is.