Jack Zhang says Airwallex turned down Stripe's roughly $1.2bn offer in 2019 because the company was too early and wanted to stay independent.
Zhang argues Stripe is payment-led while Airwallex is building toward a global business bank, which he says justified staying the course.
Airwallex has crossed $1bn in annualized revenue after ten years, with marketing now driving more than 70% of new customers.
Why Airwallex refused Stripe and built a global business bank
Airwallex is building a global bank for businesses, and Zhang believed selling to Stripe in 2019 would have cut that short. The company began as money-movement infrastructure and now earns most revenue from accounts, cards and software. That layering, plus a huge market of roughly 120m businesses, is the case for staying independent.
Takeaways
Zhang rejected Stripe's $1.2bn offer in 2019. Airwallex was about three years old and had scaled for under twelve months, going from nothing to over $1bn in volume. Zhang saw Stripe as payment-led, while Airwallex was moving toward a global business bank with more licenses.
Infrastructure shrank from 95% of revenue to 15%. Airwallex began with global money movement through APIs, then added business accounts, corporate cards and merchant acquiring. It is now building an operating system of software and AI, with software work starting in 2019.
Scale is large and still growing fast. Annualized revenue was about $900m, growing about 95% year on year, with over $200bn processed this month across about 20 markets. Airwallex has since crossed $1bn in annualized revenue after ten years.
Stablecoins offer little advantage for now. Liquidity costs about 10 to 15 basis points for exotic currencies and about a quarter of a basis point for G10 currencies. Zhang expects wider merchant and consumer adoption in three to five years, through cards and onchain systems.
Brand spending lifted growth and cut acquisition cost. Sales were mostly outbound for five years. After McLaren and Arsenal sponsorships and about four quarters of brand investment, over 70% of customers come from marketing. Last quarter new customers rose about 50% and acquisition cost fell 30%.
The US and AI drive the next phase. The San Francisco office has over 100 people, US business has grown triple digits for three years, and the office will double within twelve months. Airwallex is building AI finance agents, and Zhang says incumbent banks lack the cloud-native data to match them.
How Airwallex got started
Zhang imported coffee while working as a trading engineer at an investment bank. Banks and Western Union made international payments slow and expensive, which became the basis for Airwallex.
Lucy offered $2m for 40% within hours of meeting him, then agreed to invest $1m for 20%. The money reached his personal account three days later, before he resigned or registered the company.
Between 2015 and 2017 the product changed twice. A two-sided money marketplace never launched, and an invoicing product won a few hundred Australian small businesses, with Macquarie providing liquidity.
Tap a name to go there · tap a line to see why they are linked
On this episode 2
Show 1
An API product for large marketplaces and tuition providers helped raise the Series B. It went live at the end of 2017, but Mastercard and Tencent delivered under 1% of promised volume, forcing a search for new customers.