Collaborative apps need to create the first group, not just the first user

Where did the people brought in by ads stop?

Interview Prep.io, an AI interview-practice service built by a solo maker in Canada, spent about KRW 960,000 on Reddit ads. That was close to the cost of a low-priced laptop. The ads generated 328,300 impressions and 1,296 clicks, but not a single sign-up. The maker explained that the first step was too demanding: people had to find and upload a résumé before they could begin.

Axlura, another solo-made service from Bangladesh, saw something similar. It spent about KRW 280,000 on Google Ads and brought in 152 people. Of those, 31 ran a website check.

Inviting friends is not enough if it only increases sign-ups. After Dropbox introduced its referral programme, sign-ups rose 60% and referrals accounted for 35% of daily new sign-ups. But founder Drew Houston later recalled that 60% of referred sign-ups never turned into actual use. If friends arrive and have nothing to complete together, invitation numbers are empty numbers too.

Some Korean group services started in the reverse order. Trevari (트레바리), a book-club service, tested a reading group with 10 acquaintances, then expanded to two groups the following month. In September 2015, it launched with four clubs and 80 members. It created gatherings with defined people, books, and dates before trying to gather app users.

Frip (프립), a Korean activity-meetup service, also put offline gatherings before its app. According to its founder, it gathered about 10,000 people through travel and leisure gatherings from 2013 to 2015, recruited multiple hosts, and launched its app in 2016. Existing Frientrip users moved to Frip.

Join the broken invitation journey into one flow

Imagine a solo operator named Minsu building a weekend walking-group app. Minsu checks installs in an ad dashboard, downloads in an app-store dashboard, then exports a member list into a separate file. There are many tools, but no connected view of where one person stopped.

A person sees an ad, goes to an app store, and installs the app. After signing up by email and passing screens that ask for notification and location permissions, they see an empty screen with no friends or posts. They reopen KakaoTalk, Korea's widely used messaging app, to send an invitation link. Their friend must go through the app store and sign-up screens again. Once inside, neither person knows what they should do together.

The problem is not the invitation button's colour or the number of features. The ad said, “Find people to walk with.” Inside the app, however, the person is left alone to create an account, bring friends, create a group, and set the first schedule. The reason and context disappear between the ad and app, the app and KakaoTalk, and sign-up and the first shared action.

Reverse the order. Instead of looking for unfamiliar individuals, Minsu first finds an existing group of six to 10 people who already walk together. When the group host sets a date and goal and creates an invitation link, the link first shows “Seongsu-dong Saturday Walk,” “This Saturday, 10 a.m.,” and “5 kilometres along the Han River.” Invitees can see who they will be with and what they will do before they sign up.

After sign-up, users should land on their own group and its first action, not a general discussion board. The first action should not be finding friends. It can be marking attendance, confirming the meeting point, or logging today's walk: an action one person can start and another group member can continue. KakaoTalk does not disappear. It remains the tool for delivering invitations.

The top metric in the operator dashboard should not be installs. It should show: “people who opened the invitation link → people who joined the group → people who completed the first shared action.” Record the screen where people stopped and the time they spent there. This makes it possible to distinguish a difficult sign-up process from an empty group or a hard first activity. Looking at group-level flows before individual names reveals differences even in small experiments.

Some work still needs people. An app will struggle to replace the judgement involved in finding trustworthy hosts, actually running the first event, contacting people who do not attend, or removing disruptive participants. At the first stage of a collaborative service, it is more realistic to help one host bring in 10 people smoothly than to eliminate operations.

Elsewhere, the door opens at the group level

Meetup, which began in the United States, has hosts create groups and events before participants join. On the free Starter plan, a new host can create one group and test up to two events a month, with up to 10 people per event. To run more, the host buys a subscription. Instead of presenting an empty discussion board, the structure supplies events with dates already set.

Meetup still leaves basic discovery open to participants and charges for hosts' recurring operating tools. But if host pricing rises, gatherings themselves can disappear. Keep watching the burden on the core users who bring people together.

Discord's Student Hubs did not mix users around the world into one feed. Only students who verified a school email could enter their school's space, where they could find or share student-created chat rooms. The existing boundary of “our school” supplied both a reason to join and a topic for the first conversation.

The feature was free for students and operated in several countries, including the United States, Japan, the United Kingdom, and Australia. Discord did not disclose the number of people joining these student spaces or their activity rate, so its geographic availability and its actual usage results should be considered separately.

Strava, a US exercise-tracking service, lets paid users invite up to 199 people to create private group challenges. Group challenges for friends or clubs provide a goal, a period, and a leaderboard.

What people pay for here is not a friend list itself. It is permission to repeatedly run shared goals, rankings, and records. This is a case of a service that began with individual exercise records and expanded into a tool that connects existing friends and exercise groups.

Things you can build now

1. An invitation page for KakaoTalk groups

• What: A service that creates an invitation link containing the group name, participants, date, and first activity.
• Who: Group hosts who run running, reading, or parenting groups in KakaoTalk group chats but cannot move people to a new app.
• Why now: Showing the reason to join before asking someone to install can reduce the loss of context at the app store and sign-up screen.
• First screen: Put four things in order from the top: “group name, people you know, next event, join.”

2. Create the first shared action

• What: Provide activity templates that two or more people can complete immediately, such as logging a walk, recording one page of a book, or checking travel supplies.
• Who: Small app operators who have built features but whose new users' friend lists and discussion boards are empty.
• Why now: Completing one shared experience immediately after sign-up can reveal the reason for drop-off more clearly than continually adding features.
• First screen: Put “who to do it with, what to finish today, and how many people have completed it” on one screen.

3. An invitation drop-off recorder

• What: Record invitation opens, sign-up starts, completed sign-ups, and first activities in time order, then show where people stop most often.
• Who: Solo app operators who look at ad dashboards, app-store statistics, and member lists separately and guess at the cause.
• Why now: Download counts alone cannot tell you whether to fix the sign-up flow or an empty first screen.
• First screen: List the four stages for each group and show only the single biggest break point prominently.

4. A tool for running experiments with existing groups

• What: Bundle sign-up forms, invitations, schedule reminders, and attendance confirmation to run small group experiments without a finished app.
• Who: People who want to confirm real shared use before building an app for neighbourhood walks, office-worker book clubs, or beginner hiking.
• Why now: Instead of advertising after building the whole app, you can first check why a few existing groups would return to use it.
• First screen: Show the groups in the experiment, the next event, the number invited, and the number who actually attended at a glance.

Why this matters where you are

Check whether your product's first users already belong to a group with a shared schedule, goal, or existing boundary such as a school. The messaging channel, payment model, and group habits may differ in your market, but the journey can still be measured from invitation opening to a first shared action. Test that journey with one real group before treating installs or sign-ups as evidence of use.

What to check today

Send your current invitation link to one existing group host. Have that host and one group member go through “open link → sign up → first shared action” while you watch the screens for 30 minutes. If both finish within five minutes and explain in the same words why they should use it together, it is worth testing further. If not, fix the first screen where they stopped before building another feature.