Hey Startup enthusiasts,
This is Read Startup. Twice a week, we pick one company's boldest call and take it apart so you can see the machinery underneath.
- Today's case: Grab
- The move: Building a ride-hailing app around paper money in the smartphone era
- The puzzle: The company with the global brand and the biggest name packed up and left
- Your takeaway: A simple test for whether localisation can become your moat
- Read time: 5 minutes
Let's see what we got.
The retreat nobody in Silicon Valley expected
In March 2018, Uber handed its business in eight Southeast Asian countries to a rival that started as a taxi-booking app in Kuala Lumpur. After spending about $700 million in the region, Uber walked away with a 27.5% stake in Grab.
Grab kept building. In February 2026, it reported its first full year of net profit and more than 50 million monthly transacting users. The story behind that result starts with something unglamorous: a wad of banknotes.
The spec Uber shipped without
When Uber arrived in 2013, its app worked the way it did in San Francisco, with a credit card on file. Outside Singapore, credit card penetration across Southeast Asia sat in the low single digits.
Grab let riders pay cash from the start. Uber took roughly two years to begin accepting cash in parts of the region. For those two years, a huge share of potential customers simply couldn't pay for an Uber.
A product spec written for one market carried an invisible assumption, and that assumption quietly shrank Uber's addressable market everywhere else.
Cash is an operations problem
Accepting cash sounds like a checkbox in the app settings. In practice, it flips the flow of money. The driver collects the full fare, and the platform now has to recover its commission from thousands of people it doesn't employ.
Grab had solved this early. A spokesperson told TechCrunch the company had built a top-up process for drivers years before, so commissions could be collected reliably. Uber, playing catch-up, mixed card and cash trips for each driver so it could deduct its share of cash rides from their card earnings.
This is the part of localization that competitors underestimate. A payment button can be copied in a sprint. Collection systems, agent networks, and driver trust take years.
Grab also ran regular sessions teaching drivers how to use smartphones in markets like the Philippines and Vietnam. That's about as far from a global playbook as operations can get.
Two wheels, then a wallet
Grab read the streets the same way. In Jakarta, Ho Chi Minh City, and Bangkok, motorbike taxis were already part of daily life, and GrabBike put that habit inside the app in 2014. Uber's motorbike product arrived about two years later, by which point riders already had a default. By 2017, Grab claimed 95% of third-party taxi-hailing and 71% of private car-hailing in the region.
Then the cash habit opened a second door. Once people trusted Grab with their daily commute, GrabPay let them load money through ATMs, online banks, and convenience-store counters; no credit card needed. Co-founder Anthony Tan told WIRED in 2018 that his biggest competitor was cash itself.
A workaround for unbanked riders turned into the foundation of a fintech business. By the end of 2025, Grab's loan portfolio had grown to $1.18 billion.
Capital still mattered
Grab was never a scrappy underdog on a shoestring. In 2017 it raised $2 billion from Didi and SoftBank at a valuation above $6 billion. SoftBank also became Uber's largest shareholder that year, and it had every reason to stop two portfolio companies from torching each other's cash.
Uber was also fighting on several fronts at once, including China, India, and Southeast Asia, with a new CEO under pressure to cut losses before an IPO. Grab was focused on one region, with leadership on the ground.
So the honest version of this story is about efficiency. Both sides had money. Grab's local knowledge made each dollar buy more market share.
Why local detail compounds
Every adaptation Grab made left something useful behind. Cash collection forced it to build reliable ways to move money in and out of the app. Driver training built loyalty with the supply side of the marketplace. Ride and payment history produced the data needed to score borrowers who had no credit file at any bank.
Those layers stack up over time, and a newcomer has to rebuild all of them to compete. That's the practical meaning of a localization moat. A moat is any advantage that keeps rivals out, and this one sits in operations and relationships that take years to recreate, well beyond anything visible in the app.
The same test applies closer to home. Pathao, bKash, and Chaldal all grew by building around how Bangladeshis actually pay and move, from cash on delivery to motorbikes in gridlock.
Founder takeaways
If you are a founder yourself, here are some takeaways for your startup:
- Write your product spec from the street. List how your customers actually pay, travel, and communicate before you copy a model from another market.
- Solve the ugly back office. Collections, cash handling, and agent networks are hard to build, which is exactly why rivals struggle to copy them.
- Treat workarounds as future infrastructure. A fix for a local constraint today can become the rails for a new product line tomorrow.
- Pair capital with context. Funding buys time, and local insight decides how much market share that time is worth.
- Run the rebuild test. Write down what a well-funded foreign rival would need a full year to recreate. That list is your moat.
Forward this to a founder copying a Silicon Valley playbook today.
