Questions to ask at the end of an interview, by what you learn

Pick your closing questions by what you still do not know at that moment, not from a list you memorised the night before. The last five minutes of an interview are the only stretch where you choose the topic, and most candidates spend them asking what a typical day looks like — a question the job posting already answered. Use them on the three things a posting can never tell you: why the seat is open, how the work is judged, and who decides.

Sorted below by what each question is built to extract, and by what to do when the answer comes back thin. Bring five written down; ask the three still unanswered when the interviewer turns the floor over.

What should you ask if you only remember three?

These three work in almost any interview, for almost any role, and each one has a wrong answer that is genuinely useful to hear.

Why is this position open — where did the last person go?

Six months from now, what will make you say this hire went well?

What is the part of this job that people underestimate before they start?

Everything below is a longer version of one of those. They travel across wildly different work: a Product Manager and a Warehouse Supervisor are asking about different things, but both need to know why the seat emptied and who grades the result. If your interview runs over and you get one question, ask the first.

Why is this role open?

Ask this one using the actual word "turnover" if the answer is vague, even though most interview advice tells you it sounds negative. The reason is arithmetic. The Bureau of Labor Statistics counted 5.1 million total separations in July 2026: 3.1 million of them quits, 1.7 million layoffs and discharges. Roughly three in five people who left a job that month chose to leave it. Somebody sat in the chair you are interviewing for. Where they went, and whether the manager will say, is the highest-value fact available to you in that room.

The answer sorts into four kinds. The person was promoted internally, which tells you the role has somewhere to go. The team grew and this is a new headcount, which tells you the scope is unwritten and you will be defining it. The person left for another company, which is normal and worth one follow-up about what they went to do. Or the manager gets vague, which is itself the answer.

Quit rates vary enormously by sector, and knowing where your industry sits changes how much weight to put on the answer.

Quits rate by industry, July 2026

Quits as a percent of employment, seasonally adjusted. Quits are voluntary separations initiated by the employee.

Leisure and hospitality		3.4%
Trade, transportation and utilities		2.6%
Total nonfarm		1.9%
Professional and business services		1.8%
Private education and health services		1.8%
Manufacturing		1.4%
Financial activities		1.2%
Information		1%
Government		0.8%

A Warehouse Supervisor interviewing at a distribution centre is looking at trade, transportation and utilities, where 2.6 percent of the workforce quit in a single month. Churn on the floor is the normal condition of that sector, not a symptom, and a supervisor who describes constant hiring is describing the job rather than confessing to a problem. A Product Manager hearing the same thing in an information-sector company — where the quits rate was 1.0 percent that month — is hearing something much stranger, and should ask a second question about it.

Here is the limit of that chart, and it matters: JOLTS is a national, seasonally adjusted, industry-level series. It says nothing whatsoever about the team you would be joining. An information-sector employer can sit inside that 1.0 percent and still have lost four of its six product managers in a year. The number tells you what "normal" looks like around this employer. Only the hiring manager can tell you whether this team is normal.

How long do people stay in jobs like this one?

The follow-up question is how long the manager expects you to be there, and there is a public baseline for what "a while" means. In January 2024, the median tenure for all wage and salary workers was 3.9 years. It varies more by occupation than most people assume.

Source: BLS Employee Tenure, January 2024. These are medians across the whole occupation group, and the survey is taken every two years, so the most recent figures predate the current hiring market.

Read across those rows and the asymmetry in our two example roles becomes obvious. A Product Manager is filed by O*NET under Marketing Managers, a management occupation where the median stay is 5.7 years — so a manager describing an eighteen-month typical tenure is describing an outlier, and it is fair to ask why. A Warehouse Supervisor sits in transportation and material moving, at 3.2 years, where a three-year average tenure on the team is simply the sector working as it always has. Same fact, opposite meaning. Ask the question in a way that gets you the number: "how long has the team been together?" beats "is there a lot of turnover?", because the first has an answer and the second has a script.

How do you find out what the job actually is?

Titles are unreliable in a way that is easy to underestimate. O*NET lists "Product Manager" as a reported job title under Marketing Managers — the same code that covers Brand Manager, Marketing Director and Market Development Executive, with a 2025 median of $166,790 a year across the whole group. That median is not a product manager salary. It is the midpoint of a bucket that O*NET fills with ten reported job titles, from Marketing Coordinator to Business Development Director, and the only way to find out which one you are interviewing for is to ask.

Three questions, and how to read the answers

Each question is designed to surface one specific fact. The third column is what to do when the answer is thin.

	Ask it like this	A thin answer sounds like	What a thin answer means
What the role really is	What did the last person in this seat spend most of their time on?	We're looking for someone entrepreneurial who can wear many hats.	The scope is undefined, or it is three jobs
How you will be judged	Six months in, what will make you say this hire went well?	Just getting up to speed and being a team player.	No one has agreed on what the role is for
Who decides	Who else signs off before something ships?	We're pretty collaborative here.	There is a decision-maker you have not met
What the work costs	What is the part of this job people underestimate?	Honestly, it's a great team, no complaints.	Either the manager is new, or nobody says the hard part out loud

The middle column is not a trap. Plenty of good managers give a thin answer to a question they were not expecting, and the right move is one follow-up, not a conclusion. A Warehouse Supervisor asking "who else signs off before something ships?" and hearing "we're pretty collaborative" should follow with a specific case: who approves overtime on a Saturday when a truck is late. That version has a name attached to the answer, and the name is the thing you wanted.

Who in the loop should you ask which question?

Most advice on closing questions assumes one interviewer. A real loop has four or five, each with a different view of the job, and asking all of them the same question is a reliable way to waste the only time you control. Route the question to the person who can actually answer it.

The pattern holds across both example roles, with different names on the boxes. A Product Manager's loop is usually a recruiter, the hiring manager, two peers and a skip-level, so all five rows are live. A Warehouse Supervisor is more often interviewed by the site manager and one operations lead, which collapses the table to three rows and makes the peer question the valuable one — the operations lead is the person who knows what the last supervisor was actually doing at six in the morning. When the same person occupies two rows, ask the manager question first. It is the one that changes the fewest hands on the way to a decision.

What should you ask about the process itself?

Save one question for logistics, and make it specific enough to be answerable. It helps to know the scale of the month your interview is sitting inside.

The month your interview sits inside

7.3MJob openings, July 2026

5.1MHires, July 2026

3.1MQuits, July 2026

1.7MLayoffs and discharges, July 2026

There were 7.3 million open positions on the last business day of July 2026 and 5.1 million hires made during the month. Openings outnumber hires every month by design — a posting is a stock, counted on one day, and hires are a flow counted across thirty — so that gap is not a measure of how slow any particular process is. JOLTS does not publish time-to-hire at all. That is precisely why "we'll be in touch next week" needs a follow-up rather than a shrug: how many more conversations are in this process, who is in them, and what date do you expect to decide? Three facts, one question, and all three are things the interviewer already knows.

The two example roles are not being hired at the same volume, which is worth knowing before you read anything into a delay. O*NET projects 61,300 openings for 2024 to 2034 in the occupation that contains Warehouse Supervisor, against 34,300 in the category that contains Product Manager. That is a statement about how often each seat comes up nationally and nothing more — it does not tell you how many rounds this employer runs, and no public series does. Only the person across the table can, which is why the question has to be asked out loud. Once you have the date they expect to decide, a week of silence is a Tuesday rather than a signal.

How will you know if I am doing well?

Ask this of the person who would be your manager, and ask it last, because it is the question that most often changes the tone of the room. A manager who has thought about the role answers in specifics — a system stood up, a metric moved, a backlog cleared. A Product Manager who hears "we'll know it when we see it" has learned that success on this team is defined after the fact, by someone's mood. One who has not thought about it will describe a personality.

What should you not ask?

Two categories are worth avoiding, and they are not the ones usually listed. The first is anything you could have found in ninety seconds — the company's headcount, what the product does, whether the role is remote when the posting says it is. The second is anything that only has one socially acceptable answer, which is why "what's the culture like?" reliably produces the word "collaborative" and nothing else.

That leaves a short list of questions to retire:

• "What's the culture like?" — replace with: how does this team handle it when a deadline is going to be missed?
• "Is there room for growth?" — replace with: who was the last person promoted out of this team, and into what?
• "What's a typical day?" — replace with: what did last week look like for the person doing this job?
• "Do you have any concerns about my background?" — this one is defensible late in a process, and a mistake in a first interview, where it invites an objection the interviewer had not formed yet.

Notice that each replacement asks for a past event rather than a general description. People improvise generalities easily and remember events accurately, so the question about last week gets you a truer picture than the question about a typical day. A Warehouse Supervisor who asks how the team handled the last missed dock appointment will hear a real story, or will hear that there wasn't one, and both answers are worth having.

When to ask about pay

Ask the recruiter, not the hiring manager. Recruiters own the band, expect the question, and are usually authorised to state a range; hiring managers frequently cannot, and asking them puts a person you want on your side in the position of saying "I'd have to check." If a recruiter screen already gave you the range, do not spend a question re-opening it — spend it on scope, which is the thing that determines where in the band you land.

Go in with a public anchor rather than a feeling. O*NET's 2025 median for the occupation that contains Warehouse Supervisor is $62,890 a year, or $30.23 an hour; the category that contains Product Manager sits at $166,790. Those are national medians for whole occupational families, not for your city and not for this title, which is exactly why they work better as a sanity check than as a target. If a stated band sits far below the family median, the useful question is not "why so low" but "what does the top of this band look like, and who is in it?"

What to do this week

Write five questions for the next interview on your calendar and mark which fact each one is meant to extract. If two of them extract the same fact, cut one. If any of them could be answered by the job posting, cut it — that is the whole filter.

Then, after the interview, write down the answer to "why is this role open?" while you still remember the wording. Compare it against what the recruiter told you at the phone screen. When those two accounts do not match, you have learned more about the employer than any question could have told you directly.

The last thing: interviews only happen if applications go out, and the slow part of applying is retyping the same handful of fields into every form. HeroApply fills them from your resume, which leaves the preparation time for the part that actually decides the outcome.