The risk score is a short, evidence-only read on a booking, shown in Booking details when Show risk score is on for the location. It exists to answer one question before you hand over an asset: is there anything on record about this renter, their card, or their deposit that a manager should see first?
It is not a credit check, it does not use demographic data, and it does not decide anything for you. With one exception (Blocked, below), the score is information for your team, not a rule.
The one rule that shapes everything
Only evidence about the renter moves the score. Missing information counts for nothing.
Most operators do not collect ID. Many do not use waivers. Many renters are booking with you for the first time. None of those facts says anything about the person, so none of them changes the number. They appear in the assessment as "Not considered" so you can see they were looked for, and that is all.
The score can go up only when something is actually known and it is a concern (a failed deposit hold, a chargeback, a card the issuer refused). It can go down only when something is actually known and it is in the renter's favour (an identity verified by Stripe, clean rentals with you, a signed waiver).
A booking with nothing to check either way gets the verdict Not enough information. That is not a pass and not a fail.
Turning it on
Location settings → Hidden flags → Safety & Compliance → Show risk score.
While it is off, no booking at that location is scored and the Risk assessment dialog says so. Once on, a booking is scored the first time someone opens its details, re-scored automatically after any damage-deposit attempt, and otherwise refreshed hourly or when you press Refresh in the dialog.
What it looks at
Every assessment runs the same fourteen checks. Each one reports Clear, Verified (counted in the renter's favour), Concern, Blocked, or Not considered (nothing to look at).
| Check | Where it comes from | What counts |
|---|---|---|
| Identity document | Stripe Identity result, or a licence scanned into the customer record | Verified by a third party is a credit. A licence whose name, date of birth or address contradicts the booking is a concern. A failed verification is a concern. No ID is not considered. |
| Waiver | Signed waivers on the booking or the customer | Signed is a credit. Unsigned is not considered. |
| Damage deposit hold | Every hold attempt on this booking | Secured is clear. Failed with nothing held yet is a concern, growing with the number of failures. All retries exhausted is a serious concern. No deposit on the booking is not considered. |
| Cards used for the deposit | Card fingerprints and cardholder names on the attempts | One card is clear. Two different cards that both failed is a concern. Three or more with none succeeding is Blocked. A card in another person's name is noted at low weight if it secured the hold, and a concern only if it also failed. |
| Card issuer checks | The issuer's CVC and postal-code responses | A failed security code is a concern. A failed postal code is a small concern, smaller still once the issuer has approved a hold anyway. Nothing reported is not considered. |
| Card decline reasons | The issuer's reason on each decline | Stolen, lost, fraudulent or pickup card is Blocked. Insufficient funds is a collectability concern, and stops counting once a hold is secured. Generic declines are a small concern that grows only when the same reason lands on more than one card. |
| Stripe Radar rating | Radar's rating on every charge, including deposit pre-authorisations | Elevated is a concern. Highest, or a charge Radar itself blocked, is Blocked. No charge yet is not considered. |
| Chargebacks and fraud warnings | Disputes on this renter's bookings (by customer record or email) and card-network early fraud warnings on the cards used | Any of these is Blocked. None is clear. |
| Account standing | The customer's status and the platform blocklist | Banned, or matching a banned account's email, phone or IP, is Blocked. Flagged or warning (set by your team) is a concern. |
| Rental history with you | Completed rentals, no-shows and damage claims on this customer | Completed rentals with no incidents are a credit that grows with the count. Prior no-shows or a confirmed damage claim are a concern. A first booking is not considered. |
| Contact details | Phone, email and address on the customer record versus this checkout | Differences are a small concern. Agreement is clear but earns no credit, because the record was created from the same checkout. |
| Booking pattern | Account age, how far ahead the booking was made, booking value | A brand-new account booking a high-value rental less than a day out is a concern. Nothing else here moves the score. |
| Email and network | Email domain, IP address | A disposable email provider or a datacenter/VPN address is a concern. A free email address like Gmail is not. |
| Outstanding balance | The customer's account balance | Money owed to you is a concern, scaled by the amount. Store credit owed to them is not. |
What it deliberately does not use
- Name, age, gender, nationality, postcode or anything else about who the renter is.
- Whether the renter uses a free email provider.
- Whether the phone or email "matches" the account. The account was made from the checkout, so a match proves nothing.
- Missing ID, missing waiver, or a first booking.
How the number is calculated
Every assessment shows this arithmetic in the dialog under How this decision was made.
- Start at 5.
- Add the points for each concern. Deposit-related concerns (hold failures, card churn, decline reasons, insufficient funds, no collateral on a high-value rental) describe one event from several angles, so together they can add at most 70 no matter how many of them fire.
- Subtract the points for verified information. The most a renter can earn is about 40: identity verified (15), ten or more clean rentals (15), an account older than 30 days (5), a signed waiver (5).
- If nothing counted as hard evidence of fraud, the score is capped at 45. Soft signals (a differing phone number, an expired licence, an unpaid balance) can add up to Review but never to Elevated or High on their own.
- If a hard-fraud signal did fire, the verdict is at least Review, however many credits offset the number. A stolen identity comes with perfectly matching documents.
- If a Blocked signal fired, the verdict is High and the dialog says "Do not release the asset." No credit can offset a block, because a block is an instruction, not a score.
Bands
| Score | Verdict | What it means |
|---|---|---|
| Nothing to check either way | Not enough information | No evidence for or against. Not a pass, not a fail. |
| 0 to 15 | Low risk | Positive evidence, nothing of concern. |
| 16 to 40 | Review | Something worth a look before release. Often a deposit hold that has not gone through yet. |
| 41 to 65 | Elevated risk | Hard evidence of a problem. A manager should look before release. |
| 66 to 100, or any block | High risk | Do not release without a manager's review. |
Blocked
These are the only conditions that produce Blocked. Each is a fact, not an inference.
- The renter is banned, or their email, phone or IP matches a banned account.
- A chargeback has been filed on this booking, or this renter has filed one before.
- The card network issued an early fraud warning for a card on this booking.
- The issuer reported a card as stolen, lost, fraudulent or "pickup card".
- Stripe Radar rated a charge highest risk, or blocked it.
- Three or more different cards failed the damage deposit and none succeeded.
Reading the dialog
- The band and the number. The band is the verdict. The number is how far into the band the evidence goes.
- How this decision was made. One sentence, then the arithmetic, then the rules that applied (the 45 cap, the Review floor, a block).
- Areas of concern / Verified information. The individual signals that fired, with their points.
- What we looked at. All fourteen checks, including the ones marked Not considered.
- Payment identity. The same facts as the Identity check card: the name on each card that paid, the issuer checks, the wallet, Radar.
Frequently asked
Why does a first-time renter with no ID and no waiver show "Not enough information" instead of "Low risk"? Because nothing is known. Low risk is an affirmative statement backed by evidence. Not enough information is honest.
A regular customer's deposit failed once and the score went to Review. Is that right? Yes. One failure is only 8 points, but a deposit that has not gone through is something to know before the boat leaves. The verdict returns to Low as soon as the hold succeeds, and the dialog re-scores automatically after every attempt.
The card is in the renter's spouse's name. Why is it listed? It is noted at low weight (8 points) because a stranger's card is the most common route to a chargeback. When the hold on that card succeeded, that is all it is: a note. It only becomes a concern if the card also failed.
Why did a debit card failure count for less? Debit cards fail large holds at more than twice the rate of credit cards, mostly on daily limits. A run of failures that is all debit is discounted, and the label says "debit card" so your staff know to ask for a credit card.
Can I change the weights? Not per location. The weights are calibrated against outcomes across the platform and re-measured when they change. If a signal is firing wrongly for your business, tell support with the booking ID.