When I joined MakeMyTrip's international hotels team, we were doing 2,200 room nights a day. Booking.com was doing something like a million. We were playing a game we were structurally set up to lose.

The team had tried the obvious things. Better inventory. More supply-side partnerships. UI improvements. Discount campaigns. Room nights grew, but slowly — and the unit economics told a grimmer story. We were acquiring customers at a loss and not retaining them. The NPS was 22. When customers had a problem abroad, they were stranded.

I spent the first two months just talking to customers. Not surveys. Conversations.

What I heard, over and over, was a version of this: I was in Bangkok. Something went wrong with my booking. I couldn't reach anyone. I had to figure it out myself.

That's not a pricing problem. That's not an inventory problem. That's an empathy problem.

The insight nobody wanted to own

The pricing issue was real — Indian travelers were being shown rates that didn't account for how they actually planned. Indian consumers book closer to departure than the global average. They often plan with a group. They're price-sensitive in a specific way: not because they can't afford quality, but because they have a much stronger instinct for value, and they've been burned by opaque pricing before.

But when I raised the empathy problem — the customer service gap, the feeling of abandonment when something went wrong abroad — I hit a wall. Customer service wasn't the product team's problem. The product team owned the funnel. Customer service was... somewhere else.

This is the organizational problem that hides behind almost every product problem: the issue doesn't map cleanly to a single team's OKRs. Ownership gaps are where users go to suffer.

I spent six weeks building the business case for why fixing the experience of something going wrong was more valuable than any funnel optimization. The math wasn't complicated: if your NPS is 22, you're getting negative word-of-mouth from roughly a third of your customers. In a category where trust is the product — you're booking a room in a country where you have no fallback — that trust destruction is a multiplier on every other problem you have.

What we actually built

We rebuilt the international booking experience from the ground up with three things in mind.

Pricing clarity. Not cheaper — clearer. We surfaced total price earlier, stopped hiding taxes, and built smart defaults for how Indian travelers actually shop (flexible dates, group options, value tiers). This sounds obvious. It wasn't obvious enough to have been done.

In-destination support. We built a 24/7 chat system specifically designed for customers who were already abroad and in trouble. Not a generic chatbot. A real team with context about the booking, the property, and the local situation. This was genuinely new.

Trust signals that mattered to our specific users. Global review scores weren't wrong, but they weren't answering the question Indian travelers were actually asking: what's this hotel like for someone like me? We surfaced reviews from Indian travelers specifically, highlighted amenities that mattered (vegetarian food options, proximity to Indian restaurants, etc.).

The numbers

Room nights per day went from 2,200 to over 14,000. NPS went from 22 to 45. We became the leading OTA for international hotels among Indian travelers.

But the number I actually care about is the NPS, not the room nights. The room night growth happened because the NPS moved. That sequence matters. We didn't optimize for growth directly. We optimized for trust, and trust unlocked growth.

The Booking.com asymmetry never went away — they had better global inventory, better data, better technology for years. But they weren't building for Indian travelers specifically. They were building for everyone, which meant they were building for nobody in particular. We won on specificity.

The problem that nobody wanted to own turned out to be the whole problem. It usually is.