Rejected by Every Bank, This College Student Built a $100K/Month App by Reading TikTok Comments

Process

In the summer of 2025, Michael Que — a 20-year-old finance student at NYU — got rejected from every investment banking and finance internship he applied to. Discouraged, he decided to try something different: browsing Acquire.com, a marketplace for buying and selling small startups, to see what business ideas were already out there.

He found a height-prediction app listed for $20,000. He didn't buy it — but browsing it made him realize something: if someone would pay $20K for an app that predicts height, real demand had to exist behind it. He remembered that at age 13, he himself had obsessively googled "height predictor." With that memory in mind, he decided to build one himself.

What actually confirmed the demand wasn't market research — it was TikTok comment sections. Almost every video about height was flooded with teenagers volunteering: "I'm 15, my mom is this tall, my dad is this tall, can you predict mine?" Total strangers were willing to share their parents' heights publicly just to get a prediction. The demand was already sitting there in plain sight — nobody had taken it seriously.

He made a decision that felt genuinely embarrassing: telling people around him he was building a height-prediction app. Most reactions were mocking. But he turned that embarrassment into a competitive edge: people capable of actually building good products often avoid a category because it feels uncool — leaving the demand for whoever's willing to swallow their pride.

Within 48 hours of launch, he made his first sale. Over the next month, he posted 10 TikToks a day, experimenting with formats until he found one that actually worked: replying directly to comments asking for predictions, on video, with an actual height prediction. The format took off.

To confirm the format was repeatable and not a fluke, he ran it again on a second account — his girlfriend's — and got the same result. During this phase, monthly revenue reached about $3,000.

Next, he systematized the approach: DMing 50 UGC (user-generated content) creators a day, paying them cheap rates to replicate the same "reply-with-a-prediction" format. The entire creator network cost only about $3,000/month. That content pushed revenue to roughly $10,000/month.

Because his entire acquisition funnel ran through a single channel — TikTok — Michael could calculate precisely how many downloads came from every 100,000 views, telling him exactly where to put more money. Once the organic content and conversion model was proven, he layered in paid ads behind the winning creative. Revenue climbed from $10K to $30K, then to $100K/month.

Today, GoTall has surpassed 1 million downloads (roughly 700K on the App Store, 300K on Android), generated $434,000+ in cumulative sales since January 2026, costs just $0.03 per install, and converts 91% of users who hit the paywall. The product runs on a freemium model: basic features (stretch tracking, meal logging, sleep tracking) are free, while the core CDC-based prediction with weekly updates sits behind a subscription — $5.99/week or $34.99/year.

"I found the real gold once I was willing to completely let go of my ego." — Michael Que

Source: Starter Story via BigGo Finance · Startup Finder

Thinking

Insight 1: The demand signal might already be sitting in plain sight — nobody's taking it seriously.

Strangers on TikTok were literally begging other strangers to predict their height — that's more direct market research than any survey could ever be. Don't limit yourself to formal "user interviews" — look at where people publicly complain or ask for help in comments and forums. That's unmet demand hiding in plain view.

Insight 2: Turn the embarrassment nobody wants to touch into a moat.

People capable of building great products often steer clear of a category out of fear of looking uncool — which leaves room for whoever's willing to swallow their pride. As Michael put it: "I found the real gold once I was willing to completely let go of my ego." The more embarrassing a niche feels, the less competition it probably has.

Insight 3: Validate with the crudest possible repeat before you scale.

Running the same content format on his own account and then his girlfriend's confirmed it wasn't a fluke — it was a repeatable system. Before committing more resources, prove with the lowest-cost method possible that your playbook is systematic, not luck.

Insight 4: Going deep on one channel beats spreading thin across many.

By running everything through TikTok alone, Michael could calculate the conversion rate at every stage (views → downloads → paid) with precision. The narrower your channel, the cleaner your data — and the clearer you'll be on where to spend. Spreading across too many channels early just blurs the picture.

Insight 5: Scale with other people's content output, not by grinding harder yourself.

Paying 50 UGC creators cheap rates to replicate one proven template scaled faster than Michael producing more content alone ever could. Once you've validated a working template, the next move isn't working harder — it's handing the template to more people to execute.

Action

Step 1: Camp out in the comment sections of social platforms in the space you're considering.

Look for people publicly asking for help, predictions, or recommendations. This is the cheapest, most authentic demand validation you can do — more direct than any paid market research report.

Step 2: If you find a direction "nobody wants to touch because it feels embarrassing," take it seriously.

If you can set aside that discomfort and treat it as a real product opportunity, that hesitation from others is likely your window. Other people's reluctance is often exactly your opening.

Step 3: Test content formats with the crudest possible version until you find one that actually converts.

Even if it's just filming yourself replying to comments, keep iterating multiple times a day until you find a format with a clearly higher conversion rate. Don't rush to scale before you've found the "hit template."

Step 4: Once you find a working template, replicate it on a second account to confirm it's repeatable.

Don't settle for "it happened to go viral once" — run it again on a fully independent account to confirm it's a system, not a one-off event.

Step 5: Go all-in on one acquisition channel first, and nail down every conversion rate before scaling spend.

Views → downloads → paid — every step should be measurable. Only after the channel is proven should you spend more or hire people to replicate the content.

This isn't for you if: you can't handle people mocking or criticizing what you're building; your space has no observable "public asks for help" signal (comment sections, forum questions) to mine; or you can't accept a process of cheap, repeated testing before scaling and always want to jump straight to the finished product.