AwardOmniCard wins 3rd place at the NTT Startup Challenge 2026, chosen from 1,700+ startups across Asia-Pacific3rd place, NTT Startup Challenge 2026→

Home/Blog/Business Solutions

Expense Fraud: Common Patterns and the Checks That Catch Them

Common expense fraud patterns and the checks that catch them

Most expense misuse is not dramatic. It is small, repeated and easy to miss when claims are checked by hand at month-end. Knowing the usual patterns tells you where to put checks, and why a check that runs before the payment works better than one that runs after. This guide lists the five patterns finance teams meet most, what each looks like in the data, the check that catches it, and how to introduce checks without making honest staff feel watched.

Key numbers

Why manual checking misses so much

Manual review depends on a person reading each claim and remembering what they saw last week. The GBTA Foundation study above shows how often a plain expense report goes wrong: about one in five contained errors or missing information. If honest mistakes are that common, deliberate ones hide easily among them. The fix is not more reading. It is rules that apply the same way to every payment, and records that make a pattern visible.

Pattern 1: the same bill twice

A bill is claimed through a reimbursement and again through a card, or submitted by two people. It is the most common pattern and the easiest to catch when every bill is read and compared on date, vendor and amount. Our guide on duplicate payments shows how to set the comparison.

Pattern 2: splitting a payment to stay under a limit

A purchase above a limit is split into two or three smaller payments at the same merchant on the same day. A per-payment limit never sees it. A rule that totals spend by person, merchant and day does.

Pattern 3: personal spend in business categories

Fuel, meals or travel that belong to a weekend, not a work trip. Location and time rules help: fuel in a city the person does not work in, or a purchase at midnight on a holiday, is worth a question. See geo-fencing and time limits.

Pattern 4: altered or reused bills

A changed amount on a photo, or a bill from another month. Reading the bill by OCR and comparing it with the payment makes these stand out. Our note on receipt capture with OCR explains what OCR reads well and where a person should still look.

Pattern 5: mileage that does not add up

Self-reported kilometres that are round, repeated or longer than the route. Distance recorded from the trip itself removes the guesswork. See GPS-verified field trips and mileage.

How each pattern looks in the data and where to check it

Pattern What shows in the data Check that catches it When it runs
Same bill twice Same vendor, date and amount in two records Duplicate and anomaly check on read bills After payment, daily
Split payment Several payments, one merchant, one day Rule on total by person, merchant and day At payment
Personal spend Fuel or meals outside the person's city or hours Location and time limits At payment
Altered bill Bill amount differs from payment OCR read compared with payment At payment
Inflated mileage Round or repeated distances Distance from the recorded trip At claim

Where to put each check

The earlier a check runs, the less there is to recover. Use the table above to match each pattern to a check at the right moment.

Introducing checks without a culture of suspicion

Most staff follow the rules. Tell people what the rules are before they apply, set limits just above normal use for each role, and give a way to ask for more. Treat a flag as a question, not an accusation. Checks that feel like a trap push honest people to find workarounds, and workarounds are where real misuse starts.

How OmniCard fits

OmniCard checks policy before the money moves, with limits by employee, department, cost centre, category (MCC), merchant, location and time. Its Audit agent looks for duplicates and anomalies, its Receipt/OCR agent reads each bill, and OmniCard Motion records GPS-verified trips with mileage calculated from kilometres. Every AI action is logged, explainable and reversible by the finance team. Each payment carries who paid, where, what for, the bill and the policy it was checked against, which is the record an auditor asks for. OmniCard is run by Eroute Technologies Pvt. Ltd., an RBI-authorised Prepaid Payment Instrument (PPI) issuer and a direct NPCI issuer for RuPay cards, UPI (the @omni handle) and FASTag, with no sponsor bank in between. If you are comparing platforms, see our dated comparison with Happay, which uses the other company's own published information.

See expense management →

Common questions

What is the most common expense fraud?

The same bill claimed twice, followed by purchases split to stay under a limit. Both are easy to detect when every bill is read and compared with the payment.

Can software stop all misuse?

No. Software makes patterns visible early and applies limits before payment, but a finance reviewer still decides what a flag means.

Should limits be strict?

Set them just above normal use for each role, with a way to ask for more. Limits set too low push people to work around them.

Where should a small business start?

With the bill and the limit: capture every bill when the payment is made, and set a limit by category for each role. Add location, time and duplicate checks next.

Related: Duplicate payments · Audit-ready expense records