PakWheels is Pakistan's biggest car marketplace, and the instinct when we took it into the UAE and Saudi Arabia as ArabWheels was obvious: it works at home, run the same playbook. Same listing flow, same trust signals, same go-to-market sequencing. The instinct was wrong often enough, and specifically enough, that I've stopped trusting it anywhere since.
What actually changes across a border
Buying a used car in Lahore and buying one in Dubai aren't the same transaction wearing different currency. In Pakistan, trust was built socially. A seller's reputation traveled through word of mouth and repeat local dealers. In the UAE, the buyer pool is far more transient and international, so the same social trust signals didn't transfer; what mattered instead was verifiable seller history and transaction protection a stranger could check in minutes, not a reputation built over years. We hadn't launched a market, we'd launched a different trust problem wearing the same UI.
A playbook encodes what worked. It doesn't encode why it worked, and "why" is the only part that was ever going to survive the border.
The three things we actually rebuilt
- A shared platform underneath, three different fronts on top. The infrastructure could be common; the trust and discovery experience couldn't.
- Verified users and seller reputations, built for a market where most buyers had no prior relationship with most sellers, to cut fraud and spam that the Pakistan model had never needed to solve the same way.
- A product discovery experience rebuilt around different shopping behavior. Gulf buyers browsed and compared far more before ever contacting a seller, so the online store needed roughly twice the depth of product discovery the Pakistan flow had required.
Three markets ended up on one shared platform, deeper product discovery in the online store, and verified users and seller reputations built specifically to cut fraud and spam in a market with far less built-in social trust than home.
What I'd check before entering any new market now
Not "does our playbook apply here," it never fully does. Instead: which specific assumption is the playbook quietly making about this customer, and is that assumption about human behavior, or about infrastructure? Infrastructure travels. Behavior rarely does, and the parts of a playbook that feel most "obviously correct" are usually the behavioral assumptions nobody thought to write down as assumptions at all.