How Do Businesses Detect Fake Receipts?
Businesses detect fake receipts with a stack of checks: arithmetic (items must sum; tax must match the jurisdiction's actual rate), reference validation (does the merchant exist, does the receipt number fit its format, does the transaction appear in card feeds?), format forensics (fonts, layouts and codes that don't match the claimed merchant), file metadata on digital submissions, and pattern analytics (duplicates, thresholds-gaming, vendor anomalies across time). Single checks catch lazy fakes; the stack catches careful ones.
Written for reviewers, auditors and owners — the defensive side of the topic covered in what receipt fraud is.
Check 1: the arithmetic
Fabricators routinely fumble math: line items that don't sum to the subtotal, tax computed at a rate the claimed city doesn't charge (sales-tax rates are public and specific — 8.875% in NYC, not 8.5%), tips that change the total inconsistently, rounding that no register produces. Recompute everything; a receipt that fails its own arithmetic fails, full stop.
Check 2: does the reference world agree?
The merchant should exist at the claimed address with the claimed phone; the receipt number should fit the merchant's format and never repeat across submissions; the expense should appear in the corporate-card feed at the same merchant, date and amount (the statement cross-check is the workhorse); and for meals, the restaurant should have been open that Tuesday. Retailers extend this to full database verification (their machinery); employers approximate it with feeds and spot calls.
Check 3: format forensics
Real receipts have house styles: chain receipts follow chain layouts (store numbers, SKU formats, footer text), thermal receipts have thermal typography — monospaced, 32/48-character lines, specific truncations (the grocery dialect is hard to improvise). Tells: proportional fonts where monospace belongs, perfect kerning on a "register" receipt, missing the codes every real receipt of that chain carries, laser-crisp printing claimed as a pocket-aged thermal slip.
Check 4: the file itself
Digital submissions carry metadata: creation software (a "photo" authored by an image editor invites questions), creation dates postdating the claimed purchase, dimensions matching screenshots of templates rather than camera output, and EXIF absence where a phone photo should have it. None is proof alone; each is a thread to pull.
Check 5: the patterns
The strongest detection is longitudinal: duplicate receipts across reports or employees (submission-hash matching catches the same image twice), expenses clustering just under receipt thresholds (policy thresholds create visible bunching when gamed), one employee's vendors never appearing in anyone else's reports, and missing-receipt affidavits at outlier frequency. Modern expense systems automate most of this — the human's job is the follow-up conversation.
Process notes for teams
Verify before accusing: most anomalies are innocent (reprints, split payments, legitimate reconstructions properly labeled). Escalate on evidence, document the checks you ran, and design policy so honesty is cheap — clear substitutes for lost receipts remove the temptation that produces most amateur fakes.
The bottom line
Recompute the math, cross-check the feeds, respect the house formats, read the metadata, watch the patterns. Fakes fail one of the five with remarkable reliability — and the same stack, run calmly, clears the honest majority faster too.
Frequently asked questions
- What's the fastest way to check a suspicious receipt?
- Arithmetic first: do the items sum, and is the tax rate the jurisdiction's real one? Then the card feed: does a matching charge exist at that merchant, date and amount? Those two checks resolve most cases.
- Can fake receipts be detected from the file?
- Often — metadata showing editing software, creation dates after the claimed purchase, or screenshot dimensions instead of camera output all flag submissions for closer review. Absence of tells doesn't prove authenticity, though.
- What patterns indicate expense fraud?
- Duplicated receipt images across reports, amounts bunched just under receipt-required thresholds, vendors unique to one employee, and unusually frequent missing-receipt affidavits — pattern analytics catch what single-document review misses.
- What should I do if I suspect a fake receipt?
- Verify quietly first — reprints, split payments and mislabeled documents explain many anomalies. Document your checks, escalate per policy with evidence, and let HR/legal run the conversation. Accusation before verification is the expensive mistake.