GlamUp · B2C · AI makeup glow-up app
Let users skip paying by inviting 3 friends with a code
1.6M share/code requests; codes flooded TikTok comments; conversion barely dropped
What they did
What happened
In their words
Nicole Cheung - she makes $400k/mo from ai apps (while in college)
Nicole Cheung… and the actual purpose of it. I think a lot of people think onboarding is for data analytics. Um and that's like the the biggest purpose of it. Um I completely disagree. Obviously, it's good for tailoring your plans and all that stuff, but a lot of it is about priming creators to increase conversion. So, we would do a lot of like, for example, beauty goals, skin concerns, so that they think, you know, "Wow, I really need this app," you know?
Nicole CheungAnd then we'll tailor like a plan for them and whatever, so that will make them, you know, want to use the app a bit more. Afterwards, you will see, you know, different scans that you can do. And after you choose it, you either pay, or you can refer friends. I don't know why the code isn't showing up here, but basically, in order to view your results, you can scan your face. Afterwards, we hide everything uh behind a paywall, so we blur everything. Um you can either do view results or invite three friends. If you click view results, then we paywall you. But if you click invite three friends, we'll bring you to this like share your invite code, and you can redeem Uh, you can redeem it.
Nicole CheungBut basically, if you invite three friends to unlock your results, you can get it for free. You don't have to pay. Um and obviously, for the target customers for GlamUp, a lot of them are like younger kids, so anywhere between like, you know, 13 to like 25-year-olds. Um and you know, they don't want to pay for the app, so they would copy and paste this code and share it with their friends. But a lot of people, they actually share it on TikTok, …
Joseph ChoiSo, it seems like most people ended up trying to get the three referrals instead of paying. Do you feel like that hurt your conversion rate? Obviously, it makes up for it a lot because you're getting tons of people like sharing, and you get the viral k-factor loop. But you did hit a lot of revenue. You got like more than $600,000 total on this app. So, do do you have an idea of your conversion rates?
Nicole CheungUm our conversion rate during that time is low. I I don't think this is the best app I've built, but this is the first one. This is the first one I've built, and and I'm like happy about it. Conversion, I don't think it changed much because in order to get three of your friends to download the app, go through the welcoming and onboarding pages, and then to do your referral code, that, to be honest, takes a long time. Even for me to convince, you know, one of my friends to do it in school, or like, you know, if I'm just calling them for them to do it, you you still get impatient with it. So, a lot of the people that ended up sharing the code actually also ended up converting.
Joseph ChoiSo, you're getting the virality without sacrificing too much conversion rate, cuz people just like it It is kind of hard to get someone to download an app and use your code. Did Were Did you give any benefit to the receiver of the code, or are they just inputting the code because their friend sent it?
Joseph ChoiSo yeah, the the referral system is very um It's very interesting. Did you get inspiration from anywhere for that? Cuz I feel like a lot of people think of referral systems as k-factor, where they're trying to get multiple downloads from each download that they get, but it seems like the main benefit was it just made your content so much more viral. Have you seen anyone else do that in the app space?
Nicole CheungI think a lot of apps actually did it, or at least when I was building it, I was seeing some apps do that. Uh we do that for for my second app, Sprout, and my current app as well.
Joseph ChoiSo, it does seem re- like, somewhat repeatable. Like, you've done it. You did it with um GlamUp, and then you did it again with Sprout. Are you doing the same mechanic for your new apps?
Get tactics like this every Monday
New tactics from the week's founder interviews, each linked to where it was said.
More from this episode
- Post the same before/after makeup slideshow with one fixed hook, caption and audioFirst post hit 10M views; millions of views a day at peak
- Strip makeup off Pinterest photos, then add it back in-app for big contrastContrast made the slideshows go viral; founder claim
- Find a viral format on your own accounts first, then hire creators to copy itFounder went viral first on all four apps; scaled GlamUp to 30-40 creators
- Put the app name in the copy-paste referral message so TikTok search ranks itSearching 'GlamUp' on TikTok shows only their content; founder claim
- Triple the weekly price from $3 to $9Conversion rate went up despite the 3x price
- Raise monthly and annual prices 20%, then test another 20%Conversion and revenue per download both rose; second +20% in test
- When people close the paywall, show the same plans at 50% off
- Limit scans per week by plan, and offer a bigger plan when they run out
- Show a fake 'analyzing' scan with blurred results, and run the AI only after payment
- Ask for a rating mid-onboarding; make users tap 'I rated' to continueOther founders who copied it say ratings improved; softer version still works
- After each onboarding question, show a screen on how the app beats the alternativesFounder says it primes users to convert; no numbers
- Skip the annual plan for a job-search app; sell weekly and monthly via Stripe
- Show the paywall right after onboarding, before users can open the app+50% conversion vs letting users try features first
- Put every new creator through a video course with a quiz after each moduleOver half go viral within 2 weeks; 100M monthly views at 200+ active creators
- Recruit creators in influencer-run Discords and on Handshake, then vibe-check them in 15 minutesMany top creators had no experience; founder claim
- Pay creators $10-30 per post plus view bonuses that shrink per view at scaleFounder says the decreasing tiers work best; no numbers
- Give creators line-by-line scripts with the emotion and camera shot for each lineTrained creators averaged hundreds of thousands to millions of views per video
- Film a 6.7-second 'War is over' clip flipping a job tracker cell to 'Accepted'Hundreds of millions of views and many conversions (not attributable)
- Rebuild a hit men's looks app for women, launching a rough MVP in a month1M users in 6 months; ~$150K MRR peak
Tags: tiktok, invite-to-unlock, blurred-results, k-factor