Accountant interview questions: judgement beyond the ledger

Year-end at a food importer where I ran payroll accounting: two working days before our P11D submission deadline, reconciling benefits data, I found that four months of van-benefit charges for our delivery drivers had been applied against the wrong employee cohort — 60-odd employees affected in both directions, some over-reported, some under.

The order mattered, so here it is as it happened. First hour: sized it before raising it — not to sit on it, but because 'there's a problem' without a shape helps nobody. Rough quantification: roughly £38,000 of benefit value misallocated, material for the affected individuals if not for the accounts. Second hour: told the financial controller, in person, with the sizing and — important — with what I didn't yet know. Fix-then-mention was available; it always is. It's also how small errors become trust problems, because the day someone finds one you fixed silently is the day they re-check everything you've ever touched.

Then the recovery, split deliberately: the controller took the external question — whether we could still file cleanly or needed to flag a correction path with our advisors — while I rebuilt the allocation from the fleet-insurance schedule, which was the one source that mapped drivers to vans independently of the payroll system that had caused the error. Two of us cross-checked all 60-odd records against it; we filed with a day to spare, correct.

The control that came out of it: the benefits reconciliation now includes a cohort-level reasonableness check — total van benefit against fleet size, which would have caught this in month one rather than month four, in about 90 seconds a month. And the root cause got fixed at source: the starter-and-mover form that fed payroll had a free-text depot field; it became a dropdown.

What I'd want you to take from it: the error wasn't the test — the first two hours were. Sizing before raising, raising before fixing, and leaving behind a check that makes the silence impossible.

Our production director at a joinery firm was about to commit to a big contract for fitted wardrobes — flattering volume, a national housebuilder, and on the quote sheet it showed a healthy margin. My management accounts said the quote was built on a costing that was quietly lying to him, and telling him that in accounting language would have gone nowhere: he'd heard 'overhead absorption' before and filed it under finance noise.

So I translated the problem into his world before I raised it. I spent a morning on the shop floor with the works manager tracing how the wardrobe units actually moved through the factory — and the costing's flaw became physical: our overhead rate spread machine time evenly across all products, but the wardrobe units monopolised the CNC router, our one genuine bottleneck, at three times the rate of anything else we made. The quote was pricing the router like it was as abundant as bench space. In his language: every hour of wardrobes displaced an hour of staircases, and staircases earned roughly twice as much per router-hour.

The conversation that changed the decision took fifteen minutes and one page. Not the cost model — a picture of the router's week: here's its 40 hours; here's what the contract takes; here's what that squeezes out; here's both options in pounds per router-hour. His response was the right kind of annoyed: 'why has nobody ever shown me the factory like this?'

What changed: he went back to the housebuilder and renegotiated — smaller committed volume at a better rate, with peak-period flexibility on our side. We kept the client relationship, kept the staircase capacity, and the following year the margin on that account was genuinely healthy rather than cosmetically so. And the pounds-per-bottleneck-hour view became a standing page in the monthly pack — his request, not mine.

The lesson I'd state plainly: advice lands when it arrives in the operational unit the decision-maker already thinks in. He didn't need to understand absorption costing. He needed to see his router's week — and it was my job to walk the floor until I could show him it.

At a distribution subsidiary reporting monthly to an overseas parent, our warehouse-refit overspend was the quarter's uncomfortable number. The general manager — decent man, under real pressure from group — asked me whether a chunk of the refit costs could 'sit in prepayments for a couple of months, since we'll get the benefit later anyway'. Said casually, almost as a technical question. Those are the ones that matter, because the casual framing is how lines get moved without anyone deciding to move them.

What I did first was take the question seriously as accounting, because occasionally there IS a legitimate answer in there: I checked whether any element genuinely met the capitalisation or prepayment tests. Some did — about £11,000 of the £74,000 related to a maintenance contract paid in advance, and I moved that with a clear conscience and showed him the working. The remaining £63,000 was repairs expenditure, incurred, this period's cost. No reading of the standards got it anywhere else.

Then the conversation, and I kept it factual and unheroic: 'The £63,000 is this quarter's cost — I can't report it as anything else, and you wouldn't want my numbers on the day group audit looks at them if I could. What I CAN do is present it properly: one page for group showing the overspend, its cause, and the six-month payback on the racking changes.' I wrote that page that afternoon. Group's response, through the GM: a grumble about the quarter and a question about whether the payback could be tracked — which we then did, monthly.

The part worth naming for this role: my line if he'd pushed was already clear in my head — a direct conversation with the financial controller at group, and I'd have told the GM to his face that's where it was going next. It never needed saying, because the first no was calm enough to be final. Integrity conversations escalate badly when the first response is either outrage or wobble; the professional version is boring, specific, and comes with an alternative that solves the person's actual problem — which was never the accounting. It was standing in front of group with a bad number and no story.

The honest engine of the move: in practice I visit businesses; I want to live in one. My portfolio year gave me eleven clients, and the pattern I noticed in myself is that I was always most engaged in the weeks INSIDE a client — the glazing group I did year-end for twice is actually what started this application, because I kept thinking about their work-in-progress accounting after the engagement closed, and practice doesn't pay you to keep thinking about one client's WIP in March.

What practice gave me that industry will use: range — I've seen eleven different month-end architectures, so I know what good looks like and what expensive-but-normal looks like; audit instincts — I reconcile to independent sources by reflex; and deadline hygiene that January teaches you permanently.

What I'll have to unlearn, and I've thought about this concretely. First: the engagement rhythm. Practice work ENDS — you sign, you file, you move on. Industry is the same company, deeper, forever; the reward structure shifts from completion to improvement, and I'll need new finish lines — I plan to treat each close cycle's improvement as one. Second: the advisor's distance. In practice I could recommend and leave; in industry I'll recommend, stay, and live with the operational consequences among colleagues I see daily — which means winning people, not just being right. Third, smaller but real: practice trained me to time-record my life in six-minute units and to treat every question as billable scope; industry colleagues asking questions ARE the job, not scope creep.

And one thing I'm deliberately keeping from practice: the working-paper discipline. Files a stranger could pick up cold — because in industry the stranger is usually me, eleven months later, wondering what past-me was thinking.

Why glazing specifically: project-based revenue, staged completions, WIP judgement calls — it's the accounting I found most alive at your year-ends, and the two engagements taught me enough about the sector's rhythms to know the question I'd ask in my first month is about how retention balances get chased.