How a group-trip app got more trips out of the group chat

Bookings went from 24% to 30% of planned trips. Here's the play, and how to run it on your own product.

Group trips have a problem. Everyone's keen in the group chat. Then nobody books.

In one US survey of 2,000 women, 76% of group trips never made it out of the chat.1 If you run a travel app, that gap is your whole business.

One of our AXUM Copilot clients runs an AI app that plans group trips. Groups were starting trips in the app. Too many stalled before anyone booked. The founder was running growth on their own.

What we did

Our agents mapped every step between a group chatting about a trip and someone paying for it. That showed exactly where groups stalled.

Then the founder ran one test a week. Each test fixed one thing: activation, retention or invites. Our agents tracked the numbers and reported back every week.

The biggest lesson was about the booking itself. Every booked trip brought the whole group into the app. So a booking wasn't just revenue. It was a batch of new users.

24%to30%of planned trips booked
£7.8kmore revenue per 1,000 trips planned
A fifthmore users coming back

That's 25% more bookings. At about £130 in commission per booking, it adds up to £7.8k more revenue for every 1,000 trips planned. People came back more too: retention rose by about a fifth.

How to run this play on your product

  1. Draw the line from idea to paid. Write down every step between someone's first intent and the moment you get paid. For a trip app, that's idea in chat, trip created, friends invited, dates agreed, first booking.
  2. Find the biggest drop. Pull 30 days of numbers for each step. Fix the biggest drop first, not the one you find most interesting.
  3. Run one test a week. Pick from three buckets. Activation gets people to their first win faster. Retention brings them back. Invites get them to pull other people in. Write the bet down before you ship: "If we change X, then Y moves from A to B."
  4. Make the paid moment pull in new users. Ask who else should be in your product when someone pays. Then build the hook right there.
  5. Keep a test log. Test, result, decision. Keep what works and kill what doesn't. After a few months, that log is your growth playbook.

Steal this: the Funnel Leak Finder

Here's a simple way to do step 2 yourself. Export 30 days of user events with a user ID, event name and timestamp. Paste it into Claude or ChatGPT with this prompt.

You are a growth engineer. Below is a CSV of user events from my product for the last 30 days, with the columns user_id, event_name and timestamp.

1. Work out the main path from first visit to first payment.
2. Show how many users reach each step and the drop-off between steps.
3. Name the single biggest drop and three likely reasons for it, based only on the data.
4. Suggest three tests for next week: one for activation, one for retention and one for invites. Write each as: "If we [change], then [metric] moves from [A] to [B]."

Ask me questions if the data is unclear. Don't guess numbers.

Want us to find yours?

Finding the leak is the easy part. Fixing it every week, without dropping everything else, is where founders get stuck. That's what AXUM Copilot does, for $1,000 a month.

Or book a 20-minute call. You'll leave with your three biggest growth gaps and the plan we'd run.

— Abel

1 Talker Research for Discover Puerto Rico, online survey of 2,000 US women, June 2026.

The Growth Engine · free newsletter

Get the next play in your inbox

One growth play every week. What we tested, the real numbers and how to run it yourself.

Free. Unsubscribe anytime.