Best Practice Reviews: How Airwallex Built for Cross-Border Expansion
Expansion is usually planned as a commercial exercise. A company identifies demand in a new market, builds a route to reach it, and treats the operational apparatus underneath as something to be arranged once the commercial case is made.
Metheus and Airwallex work together on the two halves of that problem. We plan and execute market entry; Airwallex provides the regulated financial infrastructure the entry runs on, operating under financial services licences across the markets it serves. The pairing exists because the commercial decision and the operational layer beneath it are usually made by different people at different times, and the gap between them is where most of the cost sits.
The finance function is where the assumption breaks first. A business trading in one market runs one set of accounts, one currency and one view of its cash position. Add three markets and each of those becomes plural, usually without anyone having decided that it should. Local banking relationships accumulate. Conversion costs appear at every step of the operating cycle. Reconciliation absorbs the difference between systems that were never designed to talk to each other. None of this is visible in a market entry plan, and all of it arrives in the first year.
Airwallex has built a business on that gap. Founded in Melbourne in 2015, it now operates as a global financial platform for businesses, and reached an $11bn valuation in June 2026 on annualised revenue of $1.3bn. What makes it worth examining is less the growth rate than the method behind it: an expansion model built around acquiring regulatory permission market by market, and a product range that widens only as far as those permissions allow.
This review looks at how that model works, what it has produced, and which parts of it transfer to companies operating at a fraction of the scale.
Airwallex at a Glance
Airwallex operates as a global financial platform for businesses, bringing multi-currency accounts, payment acceptance, foreign exchange (FX) and transfers, billing, spend management and embedded finance products onto a single regulated infrastructure. Founded in Melbourne in 2015, it now sells to companies ranging from early-stage startups to listed enterprises, and to platforms that build financial products on top of its application programming interfaces (APIs).
The company is co-headquartered in San Francisco and Singapore, having established San Francisco as a second global headquarters in December 2025 alongside its existing Singapore base.
The current position:
Founded: Melbourne, 2015
Headquarters: San Francisco and Singapore
People: more than 2,300 employees across 27 offices
Regulatory footprint: 85 or more licences across North America, Europe, the Middle East and Asia-Pacific
Latest funding: Series H of $320m in June 2026, at an $11bn valuation
Customers: more than 676,000 businesses served, directly or through platform customers
The figure worth holding onto is the licence count. Most companies expanding internationally treat regulatory permission as an administrative cost to be absorbed on the way to a market. Airwallex has built its expansion around acquiring it, market by market, and that decision shapes almost everything else examined here, from how quickly it can enter a country to what it can sell once it arrives.
Growth That Compounds Rather Than Spikes
In March 2025, Airwallex reported $720m in annualised revenue, up 90% year on year, against $130bn in annualised transaction volume. Twelve months later the figures were $1.3bn in annualised revenue, up 74%, and $287bn in transaction volume, up more than 120%.
The growth rate fell while the absolute numbers nearly doubled. That is the shape of a business compounding rather than spiking, and it is a more useful signal than either figure alone.
The regional pattern is the more instructive detail. Growth is not distributed evenly, and the direction of the imbalance matters:
Europe, the Middle East and Africa: revenue up 116% year on year in the fourth quarter of 2025, with transaction volume up 226%
Canada: revenue up 390% over the twelve months to May 2025
Singapore: revenue up 107% across the 2025 financial year, with transaction volume up 93%
The founding Asia-Pacific region, where the company built its first decade of operations, is now growing more slowly than markets it entered later. For a business a decade old, that is the outcome an expansion programme is supposed to produce and rarely does. Most companies find the opposite: a strong home market subsidising newer territories that never reach comparable momentum.
Customer growth tells a similar story from a different angle. Airwallex served 150,000 businesses in May 2025. By mid-2026 the figure was more than 676,000 when counting both direct customers and businesses reached through platforms building on its infrastructure. The gap between those two ways of counting is itself informative: a growing share of the business now arrives through other companies' products rather than through direct acquisition.
The Cross-Border Problems That Surface During Expansion
Expansion changes a finance function before it changes anything else. A business trading in one market runs one set of accounts, one currency, one reconciliation process and one view of its cash position. Add three markets and each of those becomes plural, usually without anyone deciding that it should.
The pattern is consistent enough to describe:
Accounts fragment by geography. Each market brings a local banking relationship, opened to receive local payments, and each sits outside whatever system the finance team already uses.
Foreign exchange cost becomes structural rather than incidental. Conversion happens repeatedly across the operating cycle, on collections, on supplier payments, on repatriation, and the spread compounds at every step.
Collection and payout run on different rails. Customers pay through local methods that vary by market; suppliers and contractors expect payment through others.
Reconciliation absorbs the difference. Settlement timings, formats and failure rates differ by corridor, and the work of matching them lands on a team that has not grown at the same rate.
Financial visibility degrades last and matters most. Cash held across entities, currencies and providers is difficult to see in aggregate, which makes it difficult to deploy.
Research published by Airwallex with the Centre for Economics and Business Research in June 2026 put a figure on the aggregate cost. It estimated that roughly $330bn in working capital sits immobilised in the global financial system because of inefficiencies in cross-border business-to-business (B2B) payments, including payment failures, FX spreads, correspondent banking fees and slow settlement. Around $144bn of that sits with businesses across Europe, the Middle East and Africa, and $43.7bn in the United States.
The number is worth reading as a description of friction rather than a market sizing. What it measures is capital that exists, belongs to operating businesses, and cannot be used while it moves.
Localising a Global Platform for Local Businesses
A global platform is not global in any market where it lacks permission to operate. That constraint sits underneath most of what Airwallex has done since 2024, and it explains a pattern that looks unusual from the outside: a technology company buying small financial services firms in markets where it has no customers yet.
Buying permission rather than waiting for it
In each case the target held regulatory permissions that would otherwise have taken years to obtain.
MexPago, Mexico. An Institución de Fondos de Pago Electrónico licence, regulated by the Comisión Nacional Bancaria y de Valores, completed in 2025.
CTIN Pay, Vietnam. An Intermediary Payment Service licence issued by the State Bank of Vietnam, agreed in March 2025.
Paynuri, South Korea. Registration as an electronic financial business operator for payment gateway services and prepaid electronic payment instruments, supervised by the Financial Services Commission, together with Foreign Exchange Business registration with the Ministry of Economy and Finance. Completed in January 2026, opening the company's seventh Asian market.
Read individually these are small deals. Read together they describe a method. Rather than applying for permissions market by market and waiting through the review period, Airwallex has repeatedly bought an entity that already holds them, then built its own products behind the licence.
The acquisition is not buying revenue or technology. It is buying time, and time is the scarcest input in any expansion programme.
What follows entry is deliberately narrow
Holding a licence does not oblige a company to launch everything at once, and Airwallex has consistently done the opposite.
In South Korea the sequence was global business accounts and payment acceptance first, spend management later in the same year, with a target of around 20 local employees by the end of 2026. A defined first product, a defined second product, and a headcount ceiling.
This is the part of the approach most available to companies operating at any scale. It costs nothing to decide what you will not sell in the first year, and it protects the operating team from carrying a full product range into a market where none of it has been tested.
Capital as a signal
Commitments to individual regions follow the same logic as the licences.
In March 2026 Airwallex committed $1.135bn to the United Kingdom, Europe, the Middle East and Africa through to 2030, including around 100 senior engineering roles in London. It was the first time the company had based engineers there since entering the UK market in 2019. A separate commitment of more than $1bn covers the United States between 2026 and 2029.
Announcements of this kind are not only capital allocation. They are signals to regulators, banking partners and enterprise buyers that the company intends to stay, which carries weight in a category where changing provider is expensive and buyers price permanence into the decision.
Connecting Accounts, Payments, Foreign Exchange and Spend
The commercial argument for a single financial platform is straightforward. Every account, provider and reconciliation process a business adds while expanding creates work that does not generate revenue. Consolidating them onto one system removes that work.
The harder question is what a single platform can actually offer in any given market, and the answer is set by regulation rather than by engineering.
The range
Airwallex sells five connected product lines on shared infrastructure.
Business Accounts, Multi-currency accounts with local details, allowing collection in more than 20 currencies from over 70 countries without opening local banking relationships in each market.
Online Checkout supports 160+ local payment methods and payment collection in 130+ currencies across 180+ countries. Payment Links lets businesses accept one-time or reusable payments without building a full checkout experience. POS Payments availability and capabilities vary by market.
FX & Transfers. Payouts to more than 200 countries in over 90 currencies, with local rails used in place of correspondent banking where they exist.
Spend management. Corporate Cards, Expense Management, bill pay Purchase Orders drawing on the same balances the business collects into.
Platform APIs and embedded finance. The same capabilities offered to other companies to build financial products inside their own software
Hong Kong shows what this requires
A business in Hong Kong using the full range sees one interface. Behind it sit two separately regulated entities, each licensed by a different regulator, and each unlocking a different part of the platform.
Moving money. Airwallex (Hong Kong) Ltd is licensed as a Money Service Operator by the Hong Kong Customs and Excise Department under licence number 16-09-01929. This is what allows the core activity: multi-currency accounts, collection, FX conversion and payouts.
Earning on it. Airwallex Capital Hong Kong Limited is licensed by the Securities and Futures Commission for Type 4 advising in securities and Type 9 asset management, under central entity number BUL570. This is what permits Airwallex Yield*, where balances that would otherwise sit idle are placed into money market funds. Airwallex was the first global payments group to hold that permission in Hong Kong, and Yield launched there in June 2025 following a beta with selected local businesses.
At the customer-facing edge, the same principle applies in smaller ways. Hong Kong businesses accept payment through the Faster Payment System alongside international cards, because that is what local customers use.
Where the range fits
The sectors served share a common characteristic: revenue and costs sitting in different currencies.
E-commerce and retail. Selling into multiple markets through local payment methods, with settlement consolidated rather than fragmented by country.
Technology and software. Subscription and usage-based billing collected in customer currencies, with contractor and supplier payments going out in others.
Professional services. Client billing across borders, with expense and card management for distributed teams.
International trade. Buying in one currency and selling in several, where FX spread and settlement timing affect margin directly.
The number that validates the architecture
More than 90% of Airwallex revenue comes from customers using more than one product.
That figure matters more than any growth rate in this review. It indicates that expansion is happening inside existing accounts rather than depending on continuous new acquisition, which is the harder and more durable version of growth. It also explains why the regulatory work is worth doing. A platform that can only move money competes on price. A platform that can move, hold, spend and invest becomes the system a finance team runs on, and that is a considerably more difficult position for a competitor to attack.
Building Presence in a Category That Looks the Same
Financial infrastructure is a difficult category to market. The products are technical, the buying decision is slow, the switching cost is high, and every provider makes broadly the same promise about removing friction from global payments. Differentiation on message alone is close to impossible when the messages converge.
Airwallex has approached this from an unusual direction, and the approach is more transferable than the budget behind it.
Three functions, not one
Jon Stona, who leads global marketing at the company, has described the marketing operation as three distinct functions that have to be resourced separately: product marketing, brand, and performance. His argument is that product marketing is the one most often underfunded, and that business-to-business marketing tends toward the dry and corporate because companies treat brand as decoration rather than as a commercial asset.
That diagnosis explains what follows. The sponsorships are not an attempt to look larger than the company is. They are an attempt to make a technical product feel like something a buyer already has a relationship with.
Sponsorship that doubles as deployment
Airwallex holds two global sports partnerships: McLaren Racing in Formula One, and Arsenal, signed in July 2025 as Official Finance Software Partner.
Company research into brand perception reported a 58% increase in perceived trust and a 70% increase in likelihood of considering Airwallex for payment needs among respondents who associated the company with McLaren Racing. Those figures come from Airwallex's own study rather than an independent one, and they should be read accordingly, but the direction is consistent with what sponsorship is expected to achieve in a low-familiarity category.
The more interesting detail is what happened next. From January 2026, Airwallex payments acceptance has run hospitality payments at Emirates Stadium.
That converts the arrangement from awareness spend into something structurally different. The partner is also the customer. The deployment is live, visible, and operating at scale under conditions the buyer can imagine, which is a form of proof that no campaign produces. Most sponsorship generates recognition and stops there.
Research as a proprietary frame
The strand most available to companies without a sponsorship budget is the original research programme.
In June 2026, Airwallex published work with the Centre for Economics and Business Research estimating that around $330bn in working capital sits immobilised globally because of inefficiencies in cross-border B2B payments. The company gave the finding a name, the Global Growth Tariff, and that naming is the strategically significant part.
The mechanism is worth separating from the numbers:
A named economics partner supplies credibility the company cannot generate on its own.
A coined term supplies a frame that can be cited, argued with and repeated by people who have never used the product.
The frame describes a problem the company happens to solve, without the material functioning as an advertisement.
A comparable study with Skift surveyed 473 travel executives across seven markets, finding two thirds reporting that payment systems were eroding margin. Both pieces do the same job: they establish the problem publicly, on the company's terms, before any conversation about the product begins.
This is the part of the presence strategy that scales downward. Commissioning research is expensive, but naming a problem precisely and consistently is not, and it is available to any company willing to take a position rather than describe its features.
What This Model Asks of a Company
The approach examined here works, but it is not free of conditions. Three of them are worth stating plainly, because they determine whether any of this is available to a company reading it.
It requires capital before revenue. A licence-led strategy front-loads cost. Legal work, regulatory applications, local entities, compliance staff and safeguarding arrangements all precede the first customer, and in several markets they precede the first product. Airwallex has been able to absorb that sequence because it has raised repeatedly, most recently $320m in June 2026. A company without that backing has to either accept a slower entry or find a partner who already holds the permissions.
It requires a category with visible incumbent failure. Cross-border payments offered an unusually clear target: costs and delays that businesses could measure themselves, attached to institutions few of them felt loyalty toward. Not every market offers that. Where incumbents are performing adequately, the same operational excellence produces a much slower response.
It requires patience with narrow beginnings. The Korean entry launched two products and targeted around 20 people. That discipline only holds if the organisation is willing to look small in a new market for longer than is comfortable, and to resist launching the full range because the platform technically supports it.
The sponsorship budget belongs in this list too, though it is the least important of the four. The brand partnerships are effective, but they are the outcome of scale rather than a route to it, and a company reading this piece will get more from the research approach than from the football.
What remains transferable, once those conditions are set aside, is the sequencing logic rather than the spending. Permission first, one product, small team, expand inside the account. None of that requires a balance sheet.
What This Means for Companies Planning Expansion
The case examined here suggests a different starting question from the one most expansion programmes begin with. Not where is the demand, but what are we permitted to do once we arrive, and how long will that permission take to obtain.
Five points carry over.
1. Treat regulatory permission as an asset, not an overhead.
In regulated categories, permission determines what can be sold, when, and to whom. Costing it as an administrative line item understates its strategic weight. Mapping what each target market requires, and how long each requirement takes, belongs in the market selection process rather than in the implementation plan that follows it.
2. Treat regulatory access as part of market-entry planning.
Acquiring a small licensed local entity is not only a large-company move. In some markets it is the difference between entering this year and entering in three. The question worth asking of any target market is whether a licensed entity exists at a price that compares favourably to the cost of waiting.
3. Enter narrow on purpose.
A defined first product and a headcount ceiling protect the team from carrying an untested full range into an unfamiliar market. It also produces cleaner evidence about what is actually working, because there are fewer variables to attribute results to.
4. Measure expansion inside the account, not only at the top of the funnel.
The proportion of revenue coming from customers using more than one product tells you whether a market entry has taken root. New logos alone can conceal a business that has to keep acquiring to stand still.
5. Name the problem before selling the solution.
Establishing a frame publicly, with credible evidence behind it, does work that no product marketing can do afterwards. It is available at any budget, provided the company is willing to take a position rather than list its capabilities.
The pattern underneath all five is sequence. Permission, then presence, then product depth, then scale. Airwallex has run that order consistently across markets it entered a decade apart, and the consistency matters more than the capital. Most expansion programmes reverse it, arriving with a full product range and a marketing plan into a market where the operating conditions have not yet been established, and then spending the following two years working backwards.
Frequently Asked Questions
Why does financial infrastructure matter when entering new markets?
Because it decides how much of each sale you keep and how fast you get it. Selling into a new country means accepting the payment methods local customers actually use, holding the currency they pay in, paying suppliers and staff locally, and converting between all of it. Handled market by market, each of those becomes a separate bank relationship, a separate conversion cost and a separate reconciliation job. Most businesses discover this in the first trading year, after the market entry decision has already been made.
How can a business accept payments in a market it has just entered?
Through local payment methods rather than international cards alone, because in many markets cards are not how customers prefer to pay. Airwallex Online Checkout supports more than 160 local payment methods and collection in over 130 currencies across more than 180 countries. For businesses that do not want to build a full checkout, Payment Links accept one-off or repeat payments without development work. In-store acceptance through point of sale is available in some markets and not others, so it is worth checking country by country before planning around it.
How does a business hold and move money across several markets?
With multi-currency accounts rather than a bank account in each country. Airwallex Business Accounts provide local account details for collection in more than 20 currencies across over 70 countries, so customers pay locally while the money lands in one place. Paying out works the other way: FX and Transfers reach more than 200 countries in over 90 currencies, using local rails instead of correspondent banking where those rails exist, which is what removes the intermediary fees and the multi-day settlement wait.
Do you need a local entity or bank account to trade in a new market?
Often not, at least at the start. Multi-currency accounts with local details let a business collect payment as a local customer would expect, and pay out locally, without incorporating or opening a bank account in that country. That removes one of the slowest steps in early market entry.
A local entity becomes necessary for other reasons: employing staff directly, holding certain licences, meeting local tax registration thresholds, or selling to buyers who require a domestic counterparty in their procurement process. The practical approach is to trade first on multi-currency infrastructure, then incorporate once the market has proven itself and a specific requirement makes it unavoidable.
Why does it matter how many products a business uses on one platform?
Because every additional provider adds reconciliation work that produces no revenue. When collection, conversion, payouts and card spend run on the same balances, the finance team sees one position instead of matching several. Airwallex reports that more than 90% of its revenue comes from customers using more than one product, which reflects how the operational benefit compounds once more of the workflow sits in one place.
How can a business get set up on Airwallex when entering a new market?
Accounts can be opened directly, with no account opening fee, no monthly maintenance charge and no minimum deposit. Applications are usually reviewed within a few working days, subject to verification.
Businesses introduced through Metheus as an Airwallex partner receive priority onboarding, a dedicated account manager for the life of the account, and fees that are negotiable on request, including rebates. Which parts of the platform are available will depend on the markets involved, since product availability follows the licences held in each country.
How Metheus Can Help
Airwallex's expansion holds a lesson that applies well beyond financial services: what a company is permitted to do in a market shapes what it can sell there, and that question belongs in market selection rather than in implementation. We work with technology, SaaS and fintech companies on exactly that sequencing, from assessing which markets can realistically be entered and how quickly, to structuring entry so the first product and the first team are sized to what the market has actually proven.
Metheus is an Airwallex partner. Payment and treasury infrastructure is part of the entry groundwork rather than an afterthought to it, and we build that layer with Airwallex for clients entering new markets, so collection, conversion and payouts are running before the first local sale rather than after it. Businesses introduced through us receive priority onboarding and a dedicated account manager.