01
Travel-specific fraud profile
High average values, long booking-to-travel windows, and routes that fraudsters cluster around. A generic fraud model is the wrong tool for the job.
Fraud Shield
Catch fraud before it becomes a chargeback.
Fraud rules you can turn on, tune and own — across 3DS, card, customer, device, location and velocity signals, with every decision attached to the booking.
Rulesacross six signal categories
Tunedto your travel customer profile
Bookingevidence kept for every decision
Fraud decisions attached to the booking
Booking root
Customer payment to reporting line
£8,420Booking value
BKBooking
Booking
Travel date, lead passenger and package details
ConnectedPYPayment
Payment
Customer payment matched to booking and channel
ConnectedSTSettlement
Settlement
Acquirer settlement and fees visible
ConnectedPFProtected funds
Protected funds
Protected amount held until travel date
ConnectedATATOL
ATOL
Report line traceable back to booking movements
Connected Explainability
Fraud decisions attached to the booking
Signals, rules and outcomes sit against the booking that produced them — so chargeback representment is a query rather than an investigation.
CASHWhat has been received, settled and made available.
PROTECTEDWhat is held or restricted until travel.
REPORTINGWhat supports ATOL, APC, trust or audit reporting.
The challenge.
Travel is one of the most-targeted card-not-present categories. Bookings are high-value, the delivery window is long, the merchandise is liquid — and most fraud tools were not built for any of that.
02
Blanket rules cost real bookings
Tightening every dial blocks paying customers as well as fraud. Loosening them lets the chargeback ratio climb. Calibration is the work.
03
Generic processors leave you with the bill
When fraud lands, the chargeback is the merchant's problem — and the evidence to defend it has to be reconstructed under deadline. Most fraud rules sit too far from the booking record to help.
How Fraud Shield decides.
Each transaction runs through six categories of signal and the rules you have enabled. The outcome — allow, review or block — and the signals behind it are recorded against the booking so the decision can be defended later.
01
Gather the signals
Authentication outcome, card BIN and country, customer profile, device fingerprint and IP, billing-vs-card geography, and velocity counts across card, device, IP and email. Each signal is captured at authorisation and stays attached to the booking.
3DS Outcome captured
Location Card vs IP
Velocity Per card · IP · email
02
Apply your rules
Each rule is ID-tagged, scoped to a signal category, and configured to your tolerance — set the threshold, tune the parameter, allow the exception list. Rules cover 3DS-decline velocity, card-and-geography mismatch, proxy/VPN/Tor detection, high and medium-risk country origin, English-speaking-only allow lists, and dozens more.
Categories 3DS · Card · Customer · Device · Location · Velocity
Per rule On / Off · Parameters
Per organisation Tuned independently
03
Decision and evidence
Every decision Fraud Shield makes — and the rules and signals behind it — sits against the booking record. When a chargeback or audit query lands, the evidence is already in place rather than being reconstructed from acquirer logs.
Decision Allow · Review · Block
Reasoning Recorded
Liability shift Preserved
What you can configure.
Six signal categories, each carrying its own rule library, each rule individually switchable and tunable.
01
3DS
Decline-velocity thresholds, step-up triggers, authentication-outcome rules.
02
Card
BIN-level controls, issuer-country logic, card-type permissions.
03
Customer
Profile-based rules across new and returning customers.
04
Device
Device fingerprinting, proxy / VPN / Tor detection.
05
Location
Card-vs-IP geography, high-risk and medium-risk country lists, English-speaking allow lists.
06
Velocity
Limits per card, IP, device and email across configurable windows.
How felloh helps.
Fraud Shield runs inside the felloh payment flow rather than as a disconnected bolt-on, so every signal, rule and decision lives on the same booking-level ledger as the payment itself.
01
Fewer false positives
Rules tuned to your real customer profile, not a generic model.
02
Stronger representment
Authentication outcome and fraud signals already attached when a chargeback lands.
03
One picture, not two
Fraud, payment and reporting evidence all sit against the same booking record.
Tune Fraud Shield to your travel book.
Talk to the team about the signals, rules and exceptions that fit your customer profile.