You're working through a stack of expense claims on a Friday afternoon, reconciling them against the credit card statement before the BAS run, and one line

27/10/2026

Rule of the week: Personal Expense Patterns

The moment you notice it

You're working through a stack of expense claims on a Friday afternoon, reconciling them against the credit card statement before the BAS run, and one line makes you stop scrolling. An employee has claimed $94.20 at a supermarket, coded to "office supplies". Nothing else on the claim looks wrong — the receipt is attached, the GST is calculated correctly, the approval chain is signed off. It's just that nobody buys $94 of printer paper at Woolworths. You've seen this pattern before: a genuine mix of work purchases and personal groceries submitted together because the card was used for both and nobody separated them out before the claim went in.

This is rarely fraud. More often it's a corporate card used for the weekly shop on a day someone forgot their personal card, with every line item from that day bundled into one claim and coded without a second look. But it is still money leaving the business for someone's dinner, and it's exactly the kind of thing a reviewer should catch before it's reimbursed and booked to an expense account that overstates deductible costs.

A worked example

An employee submits a monthly expense claim with six line items totalling $612.40. Five are unremarkable — fuel, parking, a client lunch, a software subscription, a courier fee. The sixth is $78.90 at a large clothing retailer, coded to "uniforms and protective clothing". The employee's role doesn't involve a uniform. The claim is approved along with the rest because the reviewer is checking totals and GST treatment, not scrutinising each merchant name against the employee's actual duties. Multiply that by twelve claims a year, or across a team of fifteen, and it stops being a one-off.

How to find it manually

Pull the expense claims export for the period — most systems let you export by employee, merchant, and GST-exclusive amount. Sort by merchant name, not by account code, because the coding is exactly what's hiding the problem. Scan for merchant categories that sit outside the business's normal spending: supermarkets, clothing and department stores, pharmacies, entertainment venues, personal care, petrol stations where the business has no vehicles.

For each hit, check three things: does this merchant category match the employee's role, does the amount look like a single business item or a basket of mixed goods, and has this merchant appeared on this employee's claims before without challenge. A one-off coffee at a supermarket café is noise. A pattern of monthly supermarket claims coded to office supplies is not.

Be honest about the cost of doing this properly. It needs someone who knows the business well enough to judge context — a marketing contractor buying clothing for a client shoot is legitimate, a warehouse supervisor claiming the same thing is not. For a business processing a hundred or more claims a month, doing this line-by-line review every period is a genuinely slow task, and it's easy to let it slide when the BAS deadline is closer than the expense review.

FAQ

What counts as a personal expense pattern on an expense claim? Repeated or one-off claims at merchants whose typical business is personal spending — supermarkets, clothing stores, pharmacies, entertainment — coded to a business expense account without a clear business reason attached.

What's the actual red flag I should be looking for? Not the merchant alone, but the mismatch: a personal-category merchant, an employee role that doesn't explain the purchase, and a generic account code that doesn't name the item specifically.

How EXP-008 does this check

The rule works from a fixed list of merchant categories flagged as personal, so it will flag genuine business purchases made at those merchants (stationery bought at a supermarket, say) and it will miss personal spending at merchants that look plausibly business-related, like restaurants or ride-share. Severity varies because a single $12 hit reads differently from a recurring monthly pattern from the same employee. What it does reliably is surface every claim against a personal-category merchant for review, grouped by employee, so the pattern is visible without a manual export.

Finanomaly is currently waitlisted.