Travel
How Hotels Actually Get Paid for OTA Bookings
How do hotels get paid for OTA bookings? Here's a breakdown of merchant and agency models and what each one costs beyond the commission.
TL;DR
Cross-border travel bookings are declined more often than domestic ones because the issuer sees an unfamiliar foreign merchant, an unusually large amount, and a transaction pattern that does not match the cardholder's normal behavior, all at once. Local acquiring is the single largest lever, since processing domestically in the buyer's market removes the cross-border flag that suppresses approvals. Network tokens, richer authorization data, and intelligent retry logic recover most of the remainder.
A family in São Paulo tries to book your Lisbon package
They compared four operators, picked yours, chosen dates, entered passenger names, and reached the payment screen with a Brazilian credit card and every intention of spending $2,800.
0:00
/0:16
Their issuer sees a large transaction, in a foreign currency, to an acquirer it does not recognize, from a cardholder whose normal monthly spend is a fraction of that amount. It has roughly two seconds to decide, and it decides no.
Nobody at your company will ever know this happened. There is no alert, no ticket, no entry in your fraud dashboard. The family tries a second card, gets the same result, and books with the operator they had ranked second.
Multiply that by the share of your traffic that comes from outside your home market and you have the largest untracked revenue leak in most travel businesses.
Why are international bookings declined so often?
Because the issuing bank makes its decision in about two seconds, with less information than you have, and travel trips several of its caution signals simultaneously.
1. The cross-border flag itself
A transaction processed outside the cardholder's market carries a higher decline baseline before anything else is considered. Issuers apply stricter rules to foreign acquirers as a default.
2. Amount relative to normal spend
A holiday booking is often the largest single transaction a cardholder makes all year, which is exactly the profile fraud models are built to question.
3. Merchant unfamiliarity
The issuer has no history with your business. A domestic retailer the cardholder uses monthly has an approval advantage you cannot buy.
4. Merchant category coding
Travel MCCs carry risk weighting in issuer models. Coding that does not match your actual operating model can quietly cost you points of approval.
5. Thin authorization data
If your checkout sends minimal data with the authorization request, the issuer defaults to caution.
Baseline matters for context. Typical card-not-present authorization rates for US ecommerce run 85% to 90%, with tokenized wallet transactions reaching higher. Cross-border travel volume sits below that band, and most operators have never measured how far below.
What does the gap actually cost?
Work it against your own numbers rather than an industry average.
Take a platform doing $30M a year with 40% of bookings from outside its home market. That is $12M of cross-border volume. Every point of authorization rate on that segment is worth $120,000.
Move from 82% to 89% and you have recovered roughly $840,000 in bookings your existing traffic already delivered. No marketing spend produced it.
Research suggests 60% to 70% of card declines are potentially recoverable, which reframes a decline stack as deferred revenue rather than lost revenue.
What lifting approval rates looks like
One platform doubled approval rates and grew transaction volume 163% in five months after migrating.
Read the Takenos story →Which levers actually work?
Roughly in order of impact for a travel platform selling internationally.
- Local acquiring in your major source markets. Processing domestically removes the cross-border flag entirely. This is the single largest lever available and the one most platforms defer longest.
- Local payment methods where cards are not the default. In markets running on instant bank rails or wallets, a card-only checkout is not a premium experience, it is a broken one. More on that in our piece on collecting payments in markets where cards are not the default.
- Network tokens and wallet acceptance. Tokenized transactions carry stronger issuer trust signals and approve at meaningfully higher rates than raw card entry.
- Richer authorization data. Complete billing information, accurate merchant category coding, and clean descriptors all raise issuer confidence at the decision point.
- Retry logic keyed to decline codes. Retry based on the specific soft decline and issuer behavior. Blind retries erode issuer trust and make outcomes worse over time.
- Multi-acquirer routing. A transaction one path declines may approve through another. Without a second path, a soft decline is simply a lost booking.
Why issuer trust compounds, and why that favors incumbents
The uncomfortable part of cross-border acceptance is that it is partly a reputation game you start from zero.
An issuer approving a transaction is making a prediction, and predictions improve with history. A domestic retailer the cardholder uses every month has years of clean transactions behind it. Your travel brand, seen for the first time, has nothing, and issuers resolve uncertainty conservatively.
This is why sudden expansion into a new source market underperforms even when everything else is right. The demand is real, the pricing works, the localization is done, and the approval rate still sits ten points below your domestic baseline for the first several months.
It also means the fix is cumulative rather than instant. Consistent transaction data, network tokens that persist across purchases, and local acquiring all build the history that lifts approvals over time. Operators who test a market for one quarter and conclude the demand is not there are frequently measuring their own newness.
How Coinflow lifts approval rates for travel platforms
Authorization rate is not a setting you switch on. It is the output of routing, data quality, merchant categorization, retry logic, and how well the acquirer actually understands your business, which is why platforms treating it as a checkbox rarely move it.
Coinflow treats acceptance as a product surface.
Intelligent routing uses card type, issuer behavior, and merchant category coding to select the path most likely to approve, with multi-acquirer redundancy so a soft decline gets a second attempt rather than becoming a lost booking. Merchant category codes are assigned against how the business actually operates rather than defaulting to a classification issuers penalize. Local acquiring and local payment method coverage run through the same integration, so entering a new source market does not mean a new vendor.
Takenos saw rejection rates fall from 80% to low single digits after migrating, doubling approval rates and growing transaction volume 163% over five months. That was the same demand, converting.
If you have never measured approval rates by source market, talk to our team and start there.
Stop losing bookings you already won
Intelligent routing, local acquiring, and multi-acquirer redundancy in one integration.
Talk to our team →Frequently asked questions
What is a good authorization rate for cross-border travel bookings?
Use 85% to 90% as the general card-not-present baseline and expect cross-border volume to sit below it, often by five to fifteen points depending on the corridor. The more useful comparison is your own domestic rate against your own international rate, since that gap is precisely what cross-border friction is costing you. Break it down by source market rather than looking at a blended international number, because the variation between corridors is large.
Does local acquiring require a legal entity in each market?
Not necessarily. Providers with local acquiring licenses can process domestically on your behalf without you incorporating locally, though the arrangement varies by market and provider. What does change is settlement currency and reconciliation, so confirm how funds are converted and when they reach you before assuming local acquiring is purely an approval-rate decision.
Will improving approval rates increase my fraud losses?
Not automatically. Approving more legitimate transactions and approving more fraud are different outcomes, and the tools that separate them have improved considerably. The measure to watch is your fraud and dispute ratio alongside your approval rate, since optimizing either in isolation tends to damage the other. Monitoring both monthly is the practical discipline.
This content is for informational purposes only and does not constitute financial, legal, or investment advice.
