Designing group trips for the group, not the planner

01

The problem

The case asked how MakeMyTrip stays relevant when an AI can plan a trip for free. I started from my own trips with friends. In a group, one or two people end up doing all the planning, and every choice needs everyone's approval. That person keeps switching apps: reading the WhatsApp group, checking Airbnb or MakeMyTrip, coming back to say what is or isn't available, and starting again.

The survey showed how often this ends badly. Of 27 travellers aged 22 to 35, 21 had a group trip fall through before anything was booked. For 16 of those 21, the trip died on clashing dates or friends dropping out, not on choosing where to go.

02

Who the users were and what I learned

• The survey. I designed it and collected the responses: 27 Indian travellers aged 22 to 35. It confirmed that planning usually falls on one or two people. Because the sample is small, I quote counts, not percentages.
• The AI teardown. I gave 6 AI assistants (ChatGPT, Gemini, Perplexity, Meta AI, Claude and MakeMyTrip's own Myra) the same real group-trip request, and checked all 35 items they quoted against live prices and availability. The general assistants, answering from memory, got 17 of 20 items wrong. None of the 6 could book anything, and none could handle a group.

What this added up to: the hard part of a group trip is not finding options. It is getting everyone to agree and commit. No AI assistant was solving that, and it is exactly where a travel company that can actually book has the advantage.

03

Options considered

1. A MakeMyTrip bot inside a general AI assistant, such as Claude. Rejected. MakeMyTrip would become one more supplier inside someone else's product, competing on price. That turns it into a commodity.
2. A separate group-trip app. Rejected. MakeMyTrip already has Myra. A second product would split users and effort.
3. Chosen
   Group planning and booking inside Myra (chosen). It builds on what MakeMyTrip already owns: its assistant, its inventory and its ability to book. These are features added to Myra, not a new product.

04

What we designed: MMT Squad

Group-trip planning and booking inside the MakeMyTrip app.

1. Friends start a trip with Myra and share their plans and budgets up front.
2. Myra checks real availability right there, so nobody has to leave the chat to look things up.
3. The group votes on dates and stays.
4. Each person blocks their own share by UPI.
5. Once everyone is in, it books automatically. A shared trip hub then holds the itinerary and tickets.

Alongside the flow: a 5-screen user flow, the entry points for launching it inside the app, a North Star metric, and an 8-week pilot scorecard with a kill switch, so MakeMyTrip would know early if it was not working and could stop.

05

Result

• Projected ~$11M a year in net new margin for MakeMyTrip in the base case, about 5% of its operating profit. The model's range runs from $2.9M to $36.7M.
• Top 31 of 2,300+ teams, after the National Semi-Finals (115 teams).
• To be clear: this is a competition design and a projection. Nothing was built or launched.

06

What I'd do differently

27 travellers is a small sample. Before the next round I would put a clickable version of the flow in front of real friend groups planning a trip, and watch where they get stuck.