Accountant Interview Questions — Senior Level

Not "tell me your life story." They want the through-line: what kind of accounting, at what scale, on what systems.

Three beats, ninety seconds. Where you are now and what you own. The one thing you've done that maps to this job. Why this role.

Skip your GPA unless you graduated last year.

The only wrong answer is a grievance. Not because they'll disagree — half the time they'll agree with you — but because the person interviewing you is imagining how you'll describe them in two years.

Say what you're moving toward. If you were laid off, say so plainly; in this market it carries no stigma and hedging does.

Give a range you've actually researched, with the bottom of it a number you'd sign. In several US states they have to give you theirs first if you ask — so ask.

If yes, say which parts you've passed and your timeline. If no, don't apologize — say what you're doing instead, and mean it. Plenty of controllers never sat the exam.

The real one.

They are testing your politics, not your eyesight.

The failure modes are equally bad. The candidate who describes marching into a colleague's cube and correcting them publicly will not get an offer at a company with a functioning team. The candidate who says they quietly fixed it and said nothing will not get one either, because silent corrections are how material misstatements survive.

What they want: you verified it first, you went to the person before you went above them, and it got documented.

"I'm a perfectionist" is not an answer, and neither is a mistake so small it's obviously chosen for safety. Pick something that cost real money or real time.

Three parts: how it was caught, what you did in the first hour, and what changed afterward. The third part is the entire question. If your answer ends with "I double-check now," you've said nothing — double-checking is not a control.

It's worth seeing the two versions side by side, because most candidates give the first one and think it went well.

An ethics question wearing plain clothes, and it comes up more in companies that have been through a rough audit.

Don't play the whistleblower who resigns on principle, and don't play the soldier who books whatever they're told. The mature answer: you asked for the rationale, you asked for it in writing, you escalated to whoever owned the decision, and you documented what you were told. "I asked him to put the instruction in an email" is a sentence controllers love, because it's what they'd have done.

Don't say you thrive under pressure. Describe the system: what you pre-close before the month ends, how you sequence days one through five, what you cut when the calendar slips.

Usually sales, usually about missing documentation, usually about revenue timing. Keep the resolution procedural. "We agreed on a cutoff and I built the reminder into the workflow" beats any story about winning an argument.

Name the real hierarchy: statutory and payroll obligations first, then anything that blocks another person, then everything else. Say out loud that you tell people early when something will slip. That last part is what they're checking.

An honesty test. Name something real and small — repetitive data entry is the standard answer for a reason — then say what you've automated to reduce it. "I love every part of it" is not believed by anyone.

Be precise and do not inflate. Name the system, the modules you actually touched, and for how long. Someone who claims SAP fluency and then can't name the transaction screen they lived in every day loses the room instantly.

If you only know one system, say so and describe how fast you picked it up, with an example. Controllers hire for that more often than for a specific package.

The second half is the actual question, and it ambushes people who just said "advanced."

Have one function ready that you used on a real task, and say what it solved. XLOOKUP, SUMIFS, INDEX/MATCH, a pivot that replaced a manual report — any of these are good if they're true. Many employers give a short Excel test; assume you'll get one.

Segregation of duties, approval thresholds, three-way match, reconciliation sign-offs. Give one you personally operated or built, and what it caught.

They want to see you go from a number to a cause to a conversation with a non-finance person. The cause is rarely accounting — it's usually an operational decision nobody told finance about.

PBC lists, schedules that tie to the trial balance, and being the person who answers within a day. If you've been through a first-year audit or a system migration during one, lead with it.

The opening question at this level, and the one that separates people who ran a close from people who worked in one.

They want the sequence and the failure points, not the calendar. Which reconciliations you hold until last and why. What you pre-close. Where the delay actually comes from — it is almost never the ledger, it is a cost centre that will not approve an accrual. Who you chase and how early.

The tell that you have done it: you talk about other departments more than about your own entries.

An open question with no right answer and an obvious wrong one, which is listing every report you can think of.

Pick three, justify each in a sentence, and make at least one of them a balance sheet item — cash, the aged receivables profile, and the balance of whichever suspense or clearing account nobody has touched in a year is a strong set, because the third one tells you what the previous person did when they did not know where something went.

This is the whole interview compressed into one question, and there is a right answer.

You size the difference against materiality, you decide whether it is a timing item or a real one, and then — this is the part people skip — you tell the CFO before the pack goes out, not after. You post to suspense with a note, you flag the number in the pack, and you say when it will be resolved.

The wrong answers are equally common at this level. One is heroism: staying until 3am and finding it, which sounds impressive and tells the interviewer you will hide problems until they are solved. The other is holding the pack, which is a decision that belongs to the CFO, not to you.

Numbers, roles, and how long. Two AP clerks for four years is a real answer; "I mentored junior staff" is not, and everyone knows it is not.

If you have never had direct reports, say so plainly and describe what you have owned instead — a process, a subsidiary, an audit. Inflating this gets found out in the reference call.

The question senior candidates prepare for least and get judged on most.

Most answers are a coaching fable that ends with the person becoming excellent. Interviewers do not believe it, because managers know that some of these end in an exit and the ones that do are the ones worth hearing about.

What earns credit: that you named the problem to the person early and specifically, that you wrote down what "fixed" looked like, that you involved HR at the right point rather than three months late, and that you can say what you would do differently. An answer that ends with a managed exit is not a failure story — it is the one that proves you will not leave a weak reconciliation in the hands of someone who cannot do it.

Do not describe someone in a way that would identify them.

The distinction is the question. Operating a three-way match is a task; deciding there should be one, setting the threshold, and getting the ops team to live with it is the job you are interviewing for.

Give the risk you were closing, the threshold you set, why that number, what it cost the business in friction, and what it caught in the first year. If it caught nothing, say that too — a control with no hits is either well-designed deterrence or a waste of everyone's time, and knowing which is a senior conversation.

Intercompany eliminations, FX translation, and non-controlling interests — and the follow-up is usually the one about which rate you use where. Closing rate for the balance sheet, average for the income statement, and the difference lands in reserves.

If you have only worked in a single entity, say so at the top. Bluffing this ends badly because the second question is always about intercompany that does not agree.

The second clause is the question, and vagueness here is fatal because nobody who owned a forecast is vague about their variance.

Give the number, the direction, and the reason. A candidate who says "we came in 9% under on revenue because two contracts slipped a quarter, and we changed the model to phase renewals rather than assume January" has demonstrated more than any description of a budgeting process.

Not a test of whether you will stand your ground. It is a test of whether you know that this ends in a conversation with the audit partner and the CFO, and that your job is to have the paper ready — the policy, the workings, the prior-year treatment — rather than to win.

Say what your materiality argument would be and say when you would concede. Candidates who cannot describe conceding worry interviewers.

Once, in writing, with the standard cited. Then their call, and you implement it.

Name the actual mechanism — a firm's update mailing, a CPD requirement you have to meet anyway, a quarterly technical review with your auditors. "I read a lot" is not a mechanism.

Three versions of one question, and you will get whichever fits your resume: why are you not a controller yet, why would you leave a role where you already lead a team, and what do you want next.

They are all asking whether this job is a step or a parking space. The honest answer is usually about scope rather than title — a bigger close, a first consolidation, a system you have not run — and it should point at something this company actually has. Saying you want to be CFO in two years at a company with a CFO who is not leaving is a common and avoidable mistake.

You cannot answer this well without having asked them something first, which is exactly why it is asked. Use what you learned earlier in the conversation, name one thing, and be modest about it — "I'd want to see the close calendar before I promised anything, but from what you've described, the accrual cut-off is where I'd start" is stronger than a confident plan built on nothing.