Order fraud screening rules for prepaid and cash on delivery orders
Builds a simple, fair screening routine for risky orders in an online store: warning signs, a points score, and what to do at each score for prepaid and cash on delivery orders. Use it when chargebacks, fake orders or refused deliveries start eating into margins.
The output this prompt produced, unedited.
The prompt
What to give
Store and average order value: [e.g. electronics accessories, average order 1,800]
Team time for manual checks: [e.g. 30 minutes a day]
Checklist for building the rules
List warning signs that fit the data available, such as mismatched names, a new account with a high order value, many units of one resaleable item, or a history of refused deliveries.
Give each sign points, with a reason, so no single weak sign blocks an order.
Set three bands: let through, check first, hold.
For each band, give separate actions for prepaid and cash on delivery orders.
Write the check-first message to the buyer, polite and neutral.
Set a limit on manual checks that fits the team time given.
Add a weekly review of false alarms and missed fraud.
Output format
| Warning sign | Points | Why it matters | Data field |
| Band | Score range | Prepaid action | Cash on delivery action |
The verification message by WhatsApp and by email.
Weekly review steps with three numbers to track.
Guardrails
Never use religion, caste, gender, name origin or neighbourhood character as a sign; judge behaviour and order data only.
Treat signs as reasons to check, not proof; genuine buyers must have a quick way through.
Keep personal data access to the people who need it.
order fraud screeningecommerce fraud rulescod fake orderschargeback preventionorder risk scoreonline store operations
How to run this
Copy the prompt and paste it into Gemini, then set the aspect ratio to 3:4. No reference photo is needed for this one.
The prompt is written in Markdown with each part under its own heading, which is what the copy action hands over. Change one section at a time, keeping the rest intact, so you can tell which change did the work.