Business strategy
Roger Chen Tests App Ideas With Ads and Tappable Mockups
An ad can test interest before an app exists. Roger Chen explains what prototypes reveal and why Lobby still needed friends online together.
Listen to the audio analysis
Create your ownWritten for builders and creators.
Roger Chen's team filmed a version of Bro that wasn't a finished app. His cofounder Leo made a tappable mockup, then used it in a video to see whether people wanted the experience.
Editor's note
Why this matters now
Roger Chen uses 2 kinds of prototype test before committing to a consumer app. He films a mockup for ads to see whether the idea attracts interest, then taps through a more detailed prototype to find interactions that feel awkward.
In this Superwall Podcast conversation, the Lobby founder connects those early tests with the harder job of keeping a social app useful after launch. Lobby, a group video app, needed users to have friends on the service and to find them online together. For a founder planning a similar product, the distinction is between testing the appeal of an idea, testing its interactions and building the conditions for repeated use.
The source
What it says
Distilled from the original. The notes above and below are the editor's own.
An Ad Tests the Pitch Before the Full Product Exists
Chen describes using small ad campaigns to learn which product concepts and presentations attract a response. The mockup provides enough of an experience to demonstrate the idea without first building the complete application.
In work on the AI companion app Bro, his team explored interfaces such as a lock-screen presence and a floating window. Those versions gave the campaign something concrete to compare.
The result can inform both the product direction and a possible marketing format. It cannot establish that the advertised experience will retain users or that people will pay for it. A viewer responds to the promise shown in the ad.
That makes this a different test from Caleb Dean's prelaunch payments for Runify. Both can guide a build, but an ad response and a payment answer different questions about demand.
A Tappable Prototype Tests How the App Feels
Chen's team uses Proto.io for detailed, interactive mockups. The benefit he describes is being able to move through the experience: tap controls, see transitions and notice when a flow takes too long or loses its appeal.
A sequence of polished screens can still conceal a slow or awkward journey. Chen wants to experience the transitions before the team commits to building them.
A mockup can reveal an interaction problem while it is still inexpensive to change the design. It cannot test everything a working product must handle, including reliable data, real users and the conditions of a live service. It gives engineering a more specific experience to build.
Lobby Needed Friends on the First Day
Once Lobby launched, Chen's attention moved from an individual's response to the product to the relationships around that person.
He reports that in Israel, about 60% of new users had at least 5 friends on day 1, compared with roughly 20% in other markets. He also describes day-90 retention of around 25% to 30% in Israel, against roughly 12% to 15% elsewhere.
The comparison suggests that having people to use the app with was worth investigating. It does not establish that adding 5 friends caused a specific retention improvement. The account does not provide the cohort sizes or a controlled comparison separating friend connections from other differences between markets.
Chen calls this relationship density. For a group video product, the mechanism is understandable: an account with nobody familiar to call has less immediate use than one connected to an active group.
Friends Also Had to Be Online Together
A second problem remained even when users had connections. People needed to be available at the same time. Chen calls that temporal density.
Lobby tried a short, 1-minute call to make joining feel like a small commitment. A bounded interaction could be easier to enter than a video hangout with no obvious end.
The team also used group photos or boomerangs that participants could share. Chen describes these as a way for absent friends to see a gathering and want to join a later one.
These mechanisms address different obstacles. The short call lowers the commitment to joining now. The shared image makes the group activity visible to someone who missed it. Neither is explained by counting total registered users alone.
Marketing Needed a Situation the Viewer Recognized
Chen describes a hook-and-demo format: begin with a human situation, then show the app's role in it. The product demonstration gives the story somewhere to go, while the opening helps the intended audience recognize why it might care.
That principle also appears in Dan Vykhopen's LinkedIn-versus-Indeed hook example. A video can attract attention from people who are unlikely to use the product, even when the format performs well by view count.
For Bro, Chen's stated ambition was to have AI contribute to human conversations, acting as a prompt or conversational participant. That was a product direction in the interview, rather than a measured demonstration that an AI companion improves relationships.
Editor's note
What to do with this
Chen's account follows a sequence of different uncertainties. The ad asks whether the promise interests people. The tappable mockup asks whether the interaction makes sense. Lobby's retention experience asks whether users have the people and occasions they need to return.
A strong result at one stage leaves the later questions open. That is why the friend-count comparison is useful as an observation to investigate, rather than a target that guarantees retention. Find other launch experiments in the Superwall launch-and-sell playlist.
The original
How I Built the #1 App on The App Store (Twice)
The Superwall Podcast · 6 December 2025
Read next
Business strategy
Dan Vykhopen: Massive's 50,000-Email Waitlist Fell Flat
Massive’s waitlist brought few users, and one word in a video hook changed who watched. Dan Vykhopen explains the gap between attention and demand.
3 min read8 min listen
Business strategy
How NGL Turns an Instagram Reply Into Its Next User
Hunter Isaacson explains NGL’s message-sharing loop, return visits and paid tier, including what its reported scale leaves unanswered.
3 min read8 min listen