Questions to Ask Your COO
Twenty questions for the half hour you get with your own chief operating officer: a skip-level, a new-joiner introduction, or a standing mentoring slot. They assume you can say we, that you already know roughly where the work gets stuck, and that you will still be here next month to do something with the answers. If you are meeting a COO from outside your own company, the questions for a COO you have just met assume none of that.
The questions
Open any question for the note
Where does my team's work show up in what you look at?
Why ask it
Anchors a skip-level in something you can act on by Monday. Plenty of teams discover here that the number they optimise never reaches this desk at all.
Which number on our dashboards do you not trust?
Why ask it
Every operation has a metric that is quietly wrong or quietly managed, and the COO usually knows which one. Knowing it stops you building a case on a figure they discount.
What breaks first if volume goes up thirty percent?
Why ask it
Every operation has a known weak link, and COOs generally enjoy being asked about it. The named constraint is usually a person or a manual step rather than a system, which is worth remembering next time growth is announced.
Where do handoffs between teams fail most often?
Why ask it
Work that leaves one team's definition of finished and arrives at another's is where delay and rework are created. Their answer tells you which two functions have an unresolved dispute, which is usually invisible from inside either one.
Which process here still depends on one person's memory?
Why ask it
Names the single-point-of-failure risk everyone tolerates. If the answer comes quickly, this has been on their list a while, and asking what has stopped it being fixed is the productive follow-up.
What have you tried to fix twice that's still broken?
Why ask it
The second attempt is where the real obstacle shows: budget, a person, an incentive, or a system nobody will replace. This question gets at institutional resistance in a way a question about challenges does not.
How do you find out about a problem: a dashboard, an escalation, or a customer?
Why ask it
Ranks the detection systems honestly. A COO who mostly hears from customers has a monitoring problem, and one who mostly hears through escalation has a culture where bad news travels slowly.
Which metric do we track that you think we should stop tracking?
Why ask it
Invites a judgment rather than a description, and it usually names a number being gamed or a report nobody reads. It also reveals what they consider a genuine measure of the operation.
How do you tell a bad process from a bad hire?
Why ask it
One of the harder judgments in operations, and the reasoning is the valuable part. Answers that lean entirely on people tend to accompany high turnover in the same roles.
What did the last serious incident change?
Why ask it
Ask about the change rather than the incident. If nothing structural altered, the review process is theatre, and that is worth knowing before you rely on it.
What would you need to see before funding another person on my team?
Why ask it
Turns a request you were going to make anyway into a specification you can meet. Write down the evidence they name and bring back exactly that in a quarter.
What do you do when a vendor misses a commitment?
Why ask it
Reveals whether the company has leverage and uses it, or absorbs failures quietly. It is also directly useful if you manage suppliers yourself.
What's on the risk register that most people here have never thought about?
Why ask it
COOs carry a list of low-probability, high-consequence risks that rarely get discussed publicly. The answer is frequently mundane, a single supplier or an unmaintained system, which is exactly why it stays unaddressed.
Which part of the operation do you deliberately keep manual, and why?
Why ask it
Tests whether automation decisions are considered or reflexive. A good answer names volume, variation, or judgment as the reason, and it tells you where a proposal to automate would not be welcome.
How do you keep quality steady when volume spikes?
Why ask it
Ask what happened during the last spike specifically. Real answers involve pre-agreed reductions in scope; weaker ones involve people working longer, which is a plan that fails the second time.
How do you decide to kill a project?
Why ask it
Stopping things is harder than starting them, and the criteria are rarely written down. Listen for whether anything has actually been cancelled recently, since a stated process with no examples is not a process.
What do people at my level worry about that you do not?
Why ask it
The gap between the two lists is where effort gets wasted, and only someone further up can see it. Expect at least one thing your team spends real hours on to be waved away.
What's the most expensive mistake you've made in operations?
Why ask it
Senior operators usually answer this well, and the lesson is more concrete than any general advice. If the mistake is described as someone else's, you have learned something about how accountability works here.
What does a good first ninety days look like in a role like mine?
Why ask it
Gets you a specific standard from the person whose opinion carries furthest. Ask what a common early mistake is, since that half of the answer is usually more actionable.
What is one thing my team could stop doing that nobody would miss?
Why ask it
Asks for a subtraction rather than another initiative, which is rare enough that it tends to get an honest answer. If they have no answer, ask again once they have seen your team's actual workload.
Getting value from time with a COO
Practical guidance for the conversation itself
Before the meeting
Know what they own here
The title covers wildly different jobs: some COOs run all delivery, some run everything except product and engineering, some are effectively a chief of staff to the CEO. Ask a colleague what sits under them before you spend the slot finding out.
Bring three questions, not fifteen
A skip-level is usually twenty-five minutes and half of it will be their questions for you. Three good ones with follow-ups will get you further than a list you rush through.
Do the homework on your own numbers
If you work in operations, know your team's throughput, error rate and backlog before you walk in. Being able to answer a question about your own area is what turns one meeting into a continuing relationship.
Decide what you want to leave with
Context, a decision, an introduction, or a sponsor are different goals and need different questions. Pick one, because trying for all four generally gets you none.
Areas worth probing
- Constraints: which resource, system or person limits output right now, and what would relieve it.
- Boundaries: where work passes between teams, since this is where most delay and rework is created.
- Detection: how quickly problems become visible, and through which channel they arrive.
- Trade-offs: what gets sacrificed when speed, cost and quality cannot all be met.
- Single points of failure: the people, suppliers and systems with no backup.
- Stopping: what has been cancelled recently and who made that call.
- Reviews: what changed structurally after the last incident, not what was concluded.
What tends to go wrong
Asking about strategy in general
Broad questions about vision get the same answer the last five people got. Operational questions with a specific object, a process, a handoff, a metric, get you something you could not have read.
Complaining about your manager
A skip-level is not the place to escalate unless the matter is serious and you have already tried the normal route. It usually costs you more than it gains.
Pitching before you understand the constraint
An unsolicited proposal in the first meeting rarely lands, because you do not yet know what has already been tried. Ask what has been attempted, then send the proposal afterwards with that context in it.
Not following up
Send three lines within a day: what you took from it, what you will do, and one thing you would like to come back on. This is the step most people skip and the one that makes a second meeting possible.