· Jack Zhang

Jack Zhang: Airwallex — Fintech, Failure, and Growth

Airwallex's Jack Zhang built an $11B global payments infrastructure by preserving a single core insight (Swift is broken) across two failed pivots, trusting first-principles conviction over pedigreed banking experience, and treating relentless competition rather than money as the finish line.

fintechfounder-modepivothiringglobal-expansionregulationresilience0% confidence

Why this is in the corpus

A rare operator account of building a globally-regulated fintech from outside Silicon Valley — dense with pivot doctrine, hiring reversals (Citibank-pedigree hires who "knew better"), a turned-down $1.2B Stripe acquisition, and cold-outreach plays that unlocked interbank access.

Summary for skimmers

Jack Zhang lost everything at 16-17, worked 15-hour days through university in Australia, pivoted Airwallex twice while keeping the Swift-alternative insight, cold-called Macquarie for interbank access, turned down Stripe's $1.2B offer (~$375M personal), and learned first-principles thinking beats experience in hiring.

Briefing

What survives the editorial filter

This page should feel like a smart colleague already listened for you and left only the operating logic worth keeping. Not everything said in the episode makes it through.

Trust signal

Direct episode extraction

Best used for

Decision-grade retrieval metadata not yet added for this episode.

Hold lightly

No explicit downgrade reason stored yet for this episode.

Principles

Durable claims that survive beyond the speaker's biography — each with explicit limits, transferability judgment, and evidence.

Principle

A clear mission and vision is the primary magnet for exceptional talent

A clear, ambitious mission is the recruiting mechanism that pulls in amazing people.

Coming out of talent-thin Melbourne, Airwallex attracted a 12-year Palantir GM and others by the scale of the mission, not location.

Sell the mission and the exceptional people select themselves.

Principle

First-principles thinking beats domain experience when building the replacement for an incumbent system

When you are building the thing that replaces the old system, reason from first principles, not from the experience of people who ran the old system.

Airwallex's Swift-alternative required rejecting how banks 'know' cross-border payments work; the experienced hires kept steering back to the incumbent design and Jack had to re-architect their systems.

Weight the problem's logic over the resume when the resume is from the system you're replacing.

Principle

Compounding earlier gets you to the destination faster — a reason to consider consolidation

Because returns compound, starting the compounding sooner (e.g. via merger) can beat going it alone slower.

Michael Moritz argued Airwallex+Stripe would reach a trillion faster together; Jack found the compounding logic convincing even though he ultimately declined.

Time-in-market compounds; consider whether combining forces starts the clock sooner.

Principle

Higher-order purpose is the durable motivation once personal happiness is no longer the driver

When the work itself isn't fun, supporting a mission larger than yourself becomes the ultimate motivation.

Jack openly says he doesn't enjoy most of the work; what sustains him is the higher-order fulfillment of enabling millions of businesses globally.

Anchor motivation to mission impact, not to daily enjoyment.

Principle

Recruit fast and fire fast — speed on people decisions is decisive in early-stage companies

In early-stage companies, both hiring and firing must be fast — hesitation is the expensive option.

Airwallex's slow firing of the wrong hires took years and required re-doing systems; Jack names the delay itself as a top mistake.

If you're waiting to fire, you're already paying the compounding cost.

Principle

Deep regulatory and licensing complexity, once built, becomes the moat

The painful, multi-market regulatory infrastructure you build is the durable barrier competitors must also cross.

Airwallex maintains multiple licenses and continuous audits across 60+ markets — an operational burden that is also near-impossible to replicate.

The compliance burden nobody wants is the defensibility everybody lacks.

Principle

A durable wedge insight survives pivots — change the business model, keep the core problem

Pivot the go-to-market repeatedly while preserving the underlying insight and capability that made the company worth starting.

Airwallex kept its interbank/Swift-alternative wedge through two failed models before the API business took off and grew 100x.

Change what you sell, not the truth you discovered.

Principle

A society's relationship to wealth shapes a founder's risk appetite and time horizon

In a society where wealth buys little extra status, money motivates less and founders can hold out for the bigger bet.

Jack credits Melbourne's flat wealth culture for making it psychologically possible to turn down $375M and keep building.

Your locale quietly prices how tempting an exit feels.

Principle

Hire the best person the moment you find them, independent of timing or function priority

Great people are the scarce input — hire them when found, not when the org chart says it's time.

Joubin hired an out-of-sequence sales candidate against every priority argument because timing is never right and the best people don't recur on demand.

When the best person appears, the right time is now.

Principle

Adopt a professional-athlete mentality: the point is to compete against the best, not to be comfortable

Treat building as elite competition — deliberately choose to compete with the best players in the world.

Jack takes joy from competing with Stripe and Ramp specifically because they are the best, not from an easy win against weaker incumbents.

Choose your competitors like an athlete chooses opponents — the best ones.

Principle

Scarcity, not abundance, is what teaches an operator the value of money

The operator's understanding of and hunger for capital is forged by losing it, not by having it.

Jack grew up wealthy without a concept of money until his family lost everything at 16-17, when a $1.50 chocolate he couldn't afford became the moment he understood it.

A scarcity shock in a founder's history often explains the drive.

Principle

Founder intuition on a candidate should override a strong-looking CV

When your gut flags a great-on-paper candidate, trust the gut over the resume.

Jack's costliest hires looked like perfect fits on paper; the times he ignored an uneasy intuition to defer to professional advice, he was 'very, very wrong'.

Don't let a polished CV talk you out of a bad-fit feeling.

Frameworks

Reusable systems and operating models — including when they help and when they break.

Framework

Pivot-but-preserve: iterate the business model while holding the wedge insight fixed

A three-stage pivot discipline: P2P netting → SMB invoicing → API, each preserving the Swift-alternative wedge until one monetized.

Airwallex ran the same wedge through three business models; the diagnostic each time was whether the new packaging still used the interbank/Swift insight.

Keep the wedge constant and let the business model be the variable you tune.

Framework

The 'how can this person fail?' hiring test

Evaluate a candidate by asking 'how could they fail here?' and examining the worst case, rather than scoring resume fit.

Joubin hired an out-of-sequence, unconventional candidate because he couldn't construct a failure path and the downside was benign (excess demand).

Score the failure path and the worst case, not the resume.

Framework

Category positioning as a composite of known primitives: 'Ramp + Stripe with Citibank as infrastructure'

Position a novel category by summing familiar primitives the audience already understands.

Airwallex explains an otherwise-abstract global-banking vision as 'Ramp + Stripe + Citibank infrastructure' to make it instantly graspable.

'X + Y with Z underneath' makes a hard category legible fast.

Signals

What appears to be shifting, for whom it matters, and what happens if you ignore it.

Signal

Future of global banking is a vertically-integrated financial+regulatory layer with agentic finance on top

Global banking is moving to a vertically-integrated financial+regulatory infrastructure with agentic finance executing on top.

Jack contrasts today's siloed process (export file, upload to Citibank, process) with an integrated future where an agent confirms and the infra moves money and auto-reconciles.

Watch for agentic finance executing directly on integrated rails.

Signal

Geographic revenue shift: North America + Europe from 0% to ~40% in three years, US growing 3x+

Airwallex's NA/Europe revenue went 0%→~40% in three years with the US growing 3x+/year — evidence global infra scales cross-region fast.

Jack reports doubling US headcount to 500-600 this year as North America and Europe now approach 40% of revenue.

Cross-region expansion of money-movement infra can compound at 3x+.

Opportunities

Only included where there is a buyer, a real wedge, and a plausible revenue path — not vague idea theater.

Opportunity

Stagnant incumbent banks (Citibank, HSBC) are the softer competitive wedge in payments

The largest, most winnable share sits against stagnant incumbent banks, not against elite fintech startups.

Jack notes Airwallex competes with both Stripe/Ramp and Citibank/HSBC, and the incumbents make for an easier fight.

The soft, large TAM is the incumbent bank, not the sharp startup.

Opportunity

A direct central-bank-to-central-bank rail as a turnkey Swift alternative

Directly connecting national payment systems is a viable turnkey replacement for the 1970s-era Swift network.

Jack's inspection of Swift's intermediary chain revealed that direct central-bank-to-central-bank connectivity was 'actually doable'.

Old core infrastructure is an opportunity map — rebuild the rail directly.

Lessons still worth keeping

Useful takeaways that did not fully clear the bar for durable principle status.

Lesson

The P2P FX-netting model failed on simulation — it needed impossible volume to net a fraction

Airwallex's first model (P2P FX netting) was killed by a pre-build simulation showing it needed impossible volume to work.

The algorithm/simulation showed netting required billions of volume for a fraction of matches with long waits; they abandoned it and pivoted to interbank connectivity.

A cheap simulation can kill a model that would otherwise waste years.

Lesson

Firing the wrong hires took years instead of weeks — the delay was a top mistake

Airwallex should have fired its wrong hires in weeks; taking years compounded the damage and is named a big mistake.

Jack doubted himself because the hires convinced him he was wrong; the delay let the issues compound across the company.

When you know the hire is wrong, move in weeks.

Lesson

A co-founder's OFAC name-collision exposed Swift's flaw and became the founding insight

Airwallex's founding insight came from a lived operational pain: a co-founder's payments repeatedly froze due to an OFAC name collision on Swift.

Investigating why Max Lee's coffee-bean payments kept disappearing exposed Swift's intermediary-bank and 140-character-message flaws, birthing the direct-rail idea.

Investigate your own operational pains — they can be the founding insight.

Lesson

Turning down Stripe's $1.2B acquisition (~$375M personal) to keep building independently

Airwallex declined an 80x-forward-revenue $1.2B Stripe offer (~$375M to Jack) to remain independent, a decision made on happiness not economics.

Jack walked San Francisco for two weeks, verbally told Patrick yes, then reversed on returning to Melbourne when the happiness question stayed unclear.

Decide the biggest exits on what you'll feel at the end of the lockup, not the check.

Lesson

Hiring big-bank pedigree hires who 'knew better' forced years of costly system rework

Airwallex's Citibank/big-bank hires pushed the incumbent playbook, and deferring to them cost the company years of re-architecture.

The hires 'had built Swift systems' and insisted Airwallex was doing it wrong; Jack later had to re-do much of the system built under their direction.

Guard against experienced hires re-importing the incumbent design.

The Plays

Try these this week

Verb-first executable actions — each one tied to a stated outcome in the episode.

Convert a large check-writer into a co-founder so they see how the money is spent

Outcome: When a would-be investor is eager and capable, bring them in as a co-founder rather than a passive check.

Context: Within four hours of meeting, Jack offered Lucy a co-founder seat for her $1M at $5M post (20%); that stake became ~half a billion and Lucy became Airwallex's president.

“Hey Lucy mean if you wanna do it, why don't you join us as co-founder and you can see how we spend your money. Right? So we agreed she would invest a million dollar and $5 million post for 20% of the company”
— Jack Zhang
days when conviction is mutual (Lucy: ~4 hours to interest, terms next morning) per
  1. 1

  2. 2

  3. 3

  4. 4

  5. 5

  6. 6

Before you start

  • · Mutual high conviction fast
  • · A real operating role to give
  • · Willingness to share meaningful equity

Land marquee logo customers to de-risk a pivot and unlock hypergrowth

Outcome: Anchor a pivot with a small number of very large customers whose own growth carries your volume.

Context: Airwallex's API pivot signed MasterCard and Shein; transaction volume grew over 100x in 2018 as it rode those customers' scale.

“we signed up two very large customers... one of them being ambassador, which is MasterCard. And we also signed up Shein... the business growing over a hundred x in 2018 from a transaction volume point of view.”
— Jack Zhang
months to sign; ~1 year to hypergrowth (100x volume in 2018) per
  1. 1

  2. 2

  3. 3

  4. 4

  5. 5

  6. 6

Before you start

  • · A product suited to large-account needs (API money movement)
  • · Brand-name VC backing to build trust
  • · Capacity to support large customers

Write a direct pitch letter to the regulator to secure a license as an unknown startup

Outcome: Pitch the regulator directly in writing to obtain a license even when you're an unknown engineer, not an established institution.

Context: Jack, an investment-bank engineer, wrote an innovation pitch letter to Australia's ASIC; it was read and Airwallex received one of only two licenses granted that year.

“I had to write a pitch letter to the commission of asset, which is the regulator in Australia to pitch, you know, innovation illustrator that, that for some reason he actually read it and give, gimme that chance. So we, one of the two lessons they granted this that year”
— Jack Zhang
months per license; ongoing audits thereafter per
  1. 1

  2. 2

  3. 3

  4. 4

  5. 5

  6. 6

Before you start

  • · A concrete, compliant product thesis
  • · Willingness to carry ongoing audit/RFI burden
  • · One owner for compliance

Run cheap side-business experiments to surface an adjacent, bigger problem

Outcome: Operate small real experiments not for their own sake but to surface the bigger adjacent problem worth building.

Context: Jack treated the coffee/burger shops as experiments; roasting beans abroad exposed the Swift payment problem that became Airwallex.

“I mean I I never sort of thought that's kind of the business for me. And, and I, I was basically used that as experiment”
— Jack Zhang
months of operating to surface signal per
  1. 1

  2. 2

  3. 3

  4. 4

  5. 5

  6. 6

Before you start

  • · Some capital and time to run experiments
  • · Willingness to treat businesses as disposable probes
  • · Curiosity to dig into root causes

Cold-call the interbank gatekeeper at odd hours and sell the dream to win fixed connectivity

Outcome: Cold-call the target institution off-hours, convert whoever answers into an internal champion by selling the mission, and let them move management.

Context: Jack's 7:45am cold call reached an overnight-shift Macquarie trader who championed Airwallex internally, winning fixed connectivity and a $50k minimum ticket (vs tens of millions); Airwallex became Macquarie's largest FX customer.

“I cold called one day Macor bank had 7:45 AM the guy pick up the phone is a guy actually doing a overnight shift... And I pitched the idea... sold him the dream of, of building one of the largest FX and money movement platform in the world. And he believed in it and he convinced his management to, allowed us to have a fixed connectivity.”
— Jack Zhang
weeks to a few months to secure connectivity per
  1. 1

  2. 2

  3. 3

  4. 4

  5. 5

  6. 6

  7. 7

Before you start

  • · Some capital raised to signal credibility
  • · A regulatory license or clear path to one
  • · A specific, technical ask the champion can action

Raise from brand-name VCs to fund the next pivot and upgrade the team

Outcome: Raise from brand-name investors specifically to buy the credibility and runway needed to pivot and upgrade the team.

Context: Sequoia and MasterCard backing let Airwallex build a better team and funded the pivot to the API-first business that took off.

“we already raised from, you know, Sequoia and, and MasterCard and a few other famous VCs. And that is, you know, obviously that the brand kind of helped, the VC's brand helped and we were able to build a better team and, and then we have the money to pivot to this API first business.”
— Jack Zhang
quarters per
  1. 1

  2. 2

  3. 3

  4. 4

  5. 5

  6. 6

Before you start

  • · Enough traction/story to attract brand-name VCs
  • · A clear pivot thesis to fund
  • · A recruiting pipeline to convert the brand into hires

Decision Moments

Actual decisions, real outcomes

Specific decisions narrated in the episode with their outcomes and transferable lessons.

End of 2018, three years in and ~160-170 people, Airwallex received a $1.2B acquisition offer from Stripe (Sequoia-brokered, ~$800M on the cap table plus ~$375M to Jack in sign-on/retention) at 80x forward revenue — a highly aligned acquirer with Patrick Collison personally courting Jack over 20-40 conversations across 2-3 months.

Did: Jack verbally told Patrick yes in San Francisco, then reversed after returning to Melbourne, deciding on the happiness question (a 5-year lockup to age 39, unclear future pride/energy) rather than the money; co-founders left it to his call, with only Max favoring the sale.Outcome: Declined the offer and stayed independent; Airwallex later grew to ~$11B (Stripe stock also rose ~16x). Jack briefly regretted it when nearly running out of money during COVID but net stood by the decision.

Decide the biggest exits on the variable you will actually optimize (happiness/autonomy vs money), and recognize that your environment is pricing the temptation for you.

Part of an emerging decision pattern across multiple episodes

In year one the first business model (peer-to-peer FX netting) was proven unworkable by simulation (needed billions in volume to net a fraction, with long waits), and the second model (SMB invoicing) had costly acquisition and couldn't raise more money — the startup was running out of runway.

Did: Pivoted twice while preserving the core Swift-alternative/interbank wedge: killed P2P netting, tried SMB invoicing, then pivoted to an API business letting large businesses move money globally — funded by a brand-name round (Sequoia, MasterCard).Outcome: The third idea (API) succeeded: signed MasterCard and Shein, transaction volume grew over 100x in 2018, and the company scaled from 5 to ~170 people in two years.

Pivot the business model repeatedly while holding the underlying wedge insight fixed so learning and capability compound instead of resetting.

Part of an emerging decision pattern across multiple episodes

As an investment-bank engineer with a sub-10-person startup, Airwallex needed interbank FX connectivity — access normally reserved for institutions trading tens-of-millions minimum tickets — and had no relationships or standing to obtain it.

Did: Cold-called Macquarie at 7:45am; the person who answered was an overnight-shift trader with time and ambition. Jack sold him the mission of building one of the world's largest FX/money-movement platforms, and the trader championed it to management.Outcome: Macquarie granted fixed connectivity and changed its API to allow a $50,000 minimum ticket (vs tens of millions); Airwallex became Macquarie's largest FX customer and the trader later became head of FX distribution.

Off-hours cold outreach plus a vivid mission pitch can convert a random junior contact into the internal champion who unlocks gatekept institutional access.

Part of an emerging decision pattern across multiple episodes

Post-Series funding, Airwallex hired many experienced people from Citibank and big banks who insisted the company was building cross-border payments wrong and pushed the incumbent design; their pedigree and persuasiveness made Jack doubt his own first-principles read.

Did: Jack initially deferred to the experienced hires' direction, then concluded first-principles thinking mattered more than experience, re-architected much of the system, and fired the wrong hires — but only after years rather than weeks.Outcome: The wrong hires created lasting issues and forced years of rework; Jack names both the experience-first hiring and the slow firing as major mistakes.

In the domain you are disrupting, weight first-principles conviction over pedigree, and act on wrong people decisions in weeks, not years.

Part of an emerging decision pattern across multiple episodes

Tensions surfaced

Contradictions and trade-offs the episode raises — judgment calls a thoughtful operator has to navigate.

Tension

Life-changing money vs. mission and happiness in an acquisition decision

The logical, money-maximizing choice (sell to Stripe) conflicted with happiness/autonomy; Jack resolved on happiness, conceding geography shaped the call.

Logic and a co-founder said sell; the happiness question said no. Jack chose autonomy and nearly went bankrupt in COVID, yet net stood by it.

When money and mission conflict, decide on the variable you'll actually optimize — and know your context is pricing it.

Tension

Trust professional expertise vs. trust founder first-principles intuition

Experienced experts can be right and are very convincing, yet in disruption their incumbent priors are wrong — weight first-principles over pedigree.

Jack deferred to Citibank hires who convinced him he was wrong, and paid years of rework; the tension is that their persuasiveness is real even when their design is wrong for a disruptor.

Resolve the expert-vs-intuition pull by asking whether the expert's priors come from the system you're replacing.

Corpus connection

Where this episode fits for retrieval

What kinds of decisions this briefing is best pulled into.

Primary decisions

  • • hire
  • • fire
  • • strategic-bet
  • • pivot
  • • fundraise
  • • m-and-a