Six refund events. One pattern worth checking.
A fictional refund-fraud pointer for a nine-store home-and-garden chain. Tracebox links refund rows, item exposure, and store-level door counts so an investigator can decide what to verify next.
Fictional example / all records invented
Northline Supply, its locations, dates, transaction IDs, receipt lines, sensor events, SKUs, prices, and exposure are illustrative. This is an example of an investigation pointer, not a real customer record or a finding about a real person.
Signal summary / 12–18 February 2026
A repeat pattern across three stores.
Six refund transactions touch the same three SKUs in a seven-day window. The POS rows show no linked original-sale reference. That combination is a lead to verify, not a conclusion about intent.
Observed cross-store correlation
Same item mix
HX-447 and BG-210 recur in refund lines; CF-018 appears once.
Three locations
Stores 014, 027, and 031 show the pattern within the same week.
Source gap to check
These sample refund rows carry no linked original-sale reference. Reconcile the source records before drawing a conclusion.
Event timeline / UTC
The sequence is visible before the verdict.
Each point is a refund event in the fictional POS sample. The timeline keeps date, location, and amount attached to the originating transaction.
- 10:14:22 UTC
Store 014 · Cedar Row
TX-RF-20418Two 40V battery packs refunded · $299.98
- 11:06:41 UTC
Store 014 · Cedar Row
TX-RF-20463One cordless driver kit refunded · $219.99
- 17:42:03 UTC
Store 027 · Mill Crossing
TX-RF-31802Two cordless driver kits refunded · $439.98
- 15:18:27 UTC
Store 027 · Mill Crossing
TX-RF-31944One battery pack and one filter set refunded · $209.98
- 10:02:16 UTC
Store 031 · Lake Junction
TX-RF-40716Two 40V battery packs refunded · $299.98
- 16:31:49 UTC
Store 031 · Lake Junction
TX-RF-40901One cordless driver kit refunded · $219.99
Item exposure / sample values
Put a number beside the signal.
The listed exposure is the sum of the six fictional refund rows. It frames the review; it does not establish loss or recoverable value.
| SKU | Item | Qty | Unit | Exposure |
|---|---|---|---|---|
HX-447 | Cordless driver kit | 4 | $219.99 | $879.96 |
BG-210 | 40V battery pack | 5 | $149.99 | $749.95 |
CF-018 | Carbon filter set | 1 | $59.99 | $59.99 |
| Illustrative refund exposure | $1,689.90 | |||
Evidence trail / sample source rows
Every pointer keeps its source.
Transaction references resolve to receipt lines; the nearby door counts add store-flow context. Sensor counts are aggregate events, not a match to a shopper, employee, or item.
| POS reference | Receipt line(s) | Door sensor timestamps (UTC) |
|---|---|---|
TX-RF-20418 | Line 03 · BG-210 × 2 · $299.98 | 10:13:58 inbound +1 · 10:22:11 outbound +1 |
TX-RF-20463 | Line 02 · HX-447 × 1 · $219.99 | 11:04:51 inbound +1 · 11:13:09 outbound +1 |
TX-RF-31802 | Line 04 · HX-447 × 2 · $439.98 | 17:41:22 inbound +1 · 17:49:12 outbound +1 |
TX-RF-31944 | Lines 02–03 · BG-210 × 1 + CF-018 × 1 · $209.98 | 15:17:54 inbound +1 · 15:23:40 outbound +1 |
TX-RF-40716 | Line 03 · BG-210 × 2 · $299.98 | 09:58:03 inbound +1 · 10:08:02 outbound +1 |
TX-RF-40901 | Line 01 · HX-447 × 1 · $219.99 | 16:29:32 inbound +1 · 16:37:02 outbound +1 |
The sensor timestamps are fictional store-level inbound/outbound count changes near the POS events. They do not show who crossed, what they carried, or whether a refund was improper. Confirm each source record and local context during review.
Put your own signal in context.
Bring a question about refunds, voids, or shrink. We’ll map the source records and the review window a focused pilot would need.
Illustrative example only · no real retailer, person, transaction, or evidence is represented.