The Physics of Viral Consumer Apps: Nikita Bier on Latent Demand, Teen Networks, and Fighting a Hoax

Nikita's first product was Politify, a college project that calculated how presidential tax proposals would affect a user's finances and simulated the policies against zip-code tax returns. It went viral during an election season and reached about 4 million users. Governments asked for a version for their budgets, so the team spun out Outline, won a contract, joined Techstars, and had a pipeline with the Obama administration. A government shutdown canceled one contract, and Nikita realized he did not like selling software to governments; his real competency was making things go viral on the internet.

The team offered to return investor money and pivoted to consumer apps; none of the investors took the money back. Over four or five years they built many consumer apps, including Five Labs, which ingested Facebook posts and inferred personality with a model similar to Cambridge Analytica's and went viral in three days with tens of millions of profiles. They then focused on mobile and launched mapping, chat, event, and other apps, which built a muscle for understanding what people want and how to test growth. Not every app worked, but the portfolio approach taught them how to take shots and learn.

After building many app types, Nikita focused on teens. Teen habits are malleable, adults do not invite people to new apps, and teens see each other every day. He says invitations per user drop about 20% for every year of age from 13 to 18, so building for adults usually means paying for every user with ads and raising large amounts of venture capital. For a seed-stage company, that makes density and network effects extremely difficult.

Pivoting from web to mobile forced Nikita to recalibrate. He had built web apps since age 12, including an e-commerce business selling pirated games, but mobile had a much lower margin for error. He treats every tap as a miracle because users bounce to the next app quickly, and he urges teams to optimize everything in the mobile experience. The studio launched many apps, and 14 were basically duds before tbh.

Nikita cites a study that tracked how many people a person texts per year of life: it rises quickly from 14 to 18, peaks around 21, then collapses before rising again at the end of life. Contacts shrink after college and marriage, so communication products with network effects want to be on the upward curve of adding connections. That is why teens have the highest urgency to communicate and why innovation at the edges of communication often starts there.

Nikita says he learned these lessons by trying things, not by theorizing; he watched teens use apps and built for different audiences. The team got faster: their first mobile app took about a year, and their last took about two weeks. The most important instruction he gives teams is to develop a reproducible testing process, because consumer ideas are unpredictable and many shots at bat reduce risk more than any single idea. He also emphasizes empirical evidence over theories that are not grounded in watching users.

By app number 15, the company was nearly out of money and a key team member had given two weeks' notice. Nikita called a lawyer about dissolving the company, then launched the final app, tbh, in a Georgia school chosen because it had the earliest start date in the United States. Within 24 hours about 40% of the school had downloaded it, and it spread to neighboring schools as servers crashed; Nikita predicted number one in the United States in six days. Then he saw a $120,000 Amazon bill against $150,000 in the bank, raised a round, ran a competitive bidding process, and sold the company.

Nikita says the install threshold for the number one app fluctuates: it used to be roughly 80,000 to 100,000 installs, but ad-heavy companies like Temu and Threads can push it to as much as 300,000 in a day. At tbh's peak, the app was getting 360,000 installs per day. That scale explains why keeping the app online was its own crisis.

The insight for tbh came from a high-school user who described a Snapchat game where teens posted emoji prompts and friends replied with emojis; Nikita saw teens looking for a vehicle for disclosure. Then he saw Sarahah, the number one app in the United States, was entirely in Arabic—the strongest possible signal of latent demand. He defines latent demand as people going through a distortive process to obtain value; if you crystallize the real motivation, adoption can be intense. Sarahah also produced negative messages, so tbh used app-authored polls that only allowed positive feedback, anonymous to the sender but selected for the recipient.

The positive-polls mechanic produced explosive volume: one school sent 450,000 messages in its first seven days, and day-one users sent around 60 messages when a messaging app might expect three or four. The team geofenced the app to scale servers, a controversial choice because it turned off something that was working, but Nikita was confident it could relaunch and go viral again. To test apps, the team saturated a school with ads and dedicated Instagram accounts that followed students who listed their school in their bio, aiming for three exposures. Nikita stresses this was testing, not the growth strategy: it eliminated confounders and got the first 100 users, after which the app had to spread on its own. The team also ran 24/7 live chat support, which gave users white-glove service and fed product feedback from Michael Gutierrez into Slack.

Once tbh broke out, Nikita found that everything breaks and must be replaced almost every three days. The first customer-support system broke after three days, the next after seven, and that pattern repeated across the company. Geofencing the app was controversial but gave the team slack to control growth and keep the product online. This is what product-market fit looks like in consumer: not a tidy dashboard but a product that is breaking faster than the team can rebuild it.

Nikita rejects the search for a universal PMF benchmark. He cites Roger Dickey's advice: if the product is working, you will know, and if there is any uncertainty, it is not working. Consumer products are binary; people will fight to get in, and teams will invent new metrics like hourly actives per day. The bottom line is that product-market fit is obvious when you see it.

Selling tbh was draining. Facebook told Nikita it had 80 people assigned to the deal while he had one person—himself. Facebook's team visited his $1,800-a-month Oakland office and found two engineers, one designer, and Nikita, prompting the surprised question of whether this was the whole company behind the number one app in the United States. He joined Facebook's youth team, about 150 people, and became a product manager for what was effectively his first job; nine weeks after launch, tbh had sold to Facebook for over $30 million.

Nikita describes Facebook as an academic environment for social networks, where studies measured how changes affected retention and DAU. But he found that PMs were detached from design, with a separate design org owning the pixels, and he felt his own design skills atrophied over four years. In new product experimentation he built about eight standalone apps, yet it was hard to be intellectually honest about ideas like an app for teens to flirt in front of McKinsey-style reviewers. He argues that people download apps to make or save money, find a mate, or unplug, and those raw motivations are hard to frame inside a large company.

Lenny asks whether incumbents can respond, and Nikita says it is harder than founders fear. An app charts, a PM posts about it, market research studies it, a framing deck circulates, VPs review it, and development takes six to 12 months. Realistically, large companies take 12 to 24 months to respond to competitive threats. Incentives make this worse: employees need market evidence to defend a bet, so the safest path is often copying an existing product.

Nikita's hot take that product management is not real comes from this separation: PMs write documents, act as team secretaries, and chase approvals from legal, privacy, and other functions. He says data scientists look at the data and designers own the product, so PMs can be far from the product itself. He credits Snapchat and Apple, where designers run the show, with novel products, but says that model has its own problems because shipping inside a large org takes force of will and project management. He does not claim designers are a silver bullet.

Nikita also recalls being seated near Mark Zuckerberg in the new product experimentation group, which let him see how the machine runs from Zuckerberg's view. His team kept a giant pop-art painting of Tim Cook that had hung in their old office, and when they moved near Zuckerberg they put it up so the Apple CEO overlooked him. Eventually an executive assistant asked him to take it home, since a painting of another big-tech executive could not stay there.

After leaving Facebook, Nikita tweeted that he was working on a new app, and the internet turned it into a meme before he had even committed to a company. Then the market crashed, the Fed hiked rates, and his portfolio fell about 30%, which pushed him toward making money. He had always wondered what would have happened if tbh had monetized, since the number one support message was 'Can I pay to reveal who sent me polls?' He rebuilt the concept with engineer Zay Turner and tested it at the same Georgia school on the same day five years later; people sent messages, but the app was not growing.

tbh had grown partly through text invites sent with Twilio, but the regulatory environment had changed: texts can no longer be sent from a server and must come from the user's device. That meant Gas users had to open a compose window and hit send, which killed invite volume. Nikita says Gas required about nine launches, including a rename and features that never existed on tbh, and became a zero-to-one development cycle for growth in a new climate. He validated in sequence—core flow, spread within a school, hopping between schools, and payment—while staying above board. He frames the internet as a living system that retaliates when companies abuse users, so he tries to design growth systems legally and ethically.

Gas was designed to make teens feel better, and Nikita says the team received daily messages from users saying they had reconsidered suicide or self-harm. The app also ensured everyone got votes by placing users who had not been voted on recently into polls at a higher frequency. Then a human-trafficking hoax began with a Snapchat screenshot and an App Store review; Nikita saw it had been screenshotted dozens of times and warned his team it could kill the company. The hoax spread because it was attention-grabbing, and other founders told him apps had shut down after similar claims. The team fought on every vector: a Washington Post headline saying Gas was not for human trafficking, public retractions from schools and police, removal of review-bombed App Store reviews, and a TikTok from Nikita's girlfriend explaining the hoax. Deletions peaked at 3% of users per day and fell to 0.1% through relentless effort; Nikita even reached the TikTok CEO to remove videos, and he says the key is to make the hoax less viral than the app.

Gas became a blowout: about $11 million in sales, 10 million downloads, and a team running almost entirely on AWS, Mixpanel, and other startup credits with no investors. Nikita negotiated every vendor bill down and treated the app as pure cashflow. Three companies wanted to buy it, and the team ultimately sold to Discord. Naming mattered more than expected: under the name Crush, invitations dropped because boys invite boys and did not want to invite friends to a pink-icon app called Crush; the app indexed about 60–65% women, so they made the icon black with a flame, called it Gas, and invite rates jumped.

Lenny asks whether Nikita wants to build a durable consumer app and whether it is possible. Nikita says incumbents in messaging and broadcasting have large network-effect moats and have improved their products, so an order-of-magnitude better experience is non-trivial. Snapchat showed a new messaging style could break through, but such edges may not be durable. Finding durability in consumer social is a black-swan event; retention has tremendous randomness, with roughly one every decade.

Growing a product, by contrast, can be a science: Nikita says with certainty that if you are good at the job, you can make an app grow and go viral. His favorite part of product development is building through the night and watching the app take over the internet. Dropping an app in Georgia and seeing 40% of a high school near his Los Angeles home download it a week later is a profound feeling for a three-person team.

At Facebook, many peers looked up to VPs and wanted large organizations and reports. Nikita found that VPs were sometimes jealous of him because he had built something high-impact, made money, and did not have to become a people manager for years. He may one day run a venture-scale business, but he likes his current quality of life and does not think running a large corporation is fun.

Lenny notes that Nikita does hope to build a venture-funded business one day, and Nikita says he would keep the team lean. He points to Elon Musk as an iconic founder who keeps teams thin and stays in the weeds on product. He also argues that after seven rounds of dilution, some IPO proceeds are comparable to what he can earn from 90 days of work on his own apps.

Nikita sees iOS 18's contact-permission change as a major threat to consumer app growth. Apps average about 65% approval for contacts access, but the new flow asks users to select specific contacts from an alphabetical list. Many users have hundreds of contacts, and the first entries may be punctuation symbols; even if a user adds a few friends, those friends may not join, and there is no over-the-top graph to connect them. He expects contact sync to become much harder, entrench incumbents, and force founders to find a plan B; he also criticizes the flow as badly designed for privacy and usability.

Nikita says 2024 attention spans are about three seconds, so if an app cannot demonstrate value immediately, it is over. With Dupe, he saw Deal Hop—paste a product page and find the cheapest version online—and argued it should be the whole company. They needed a memorable way to expose the aha moment, so they bought Dupe.com and told users to type it in front of any product URL. Videos of the trick went viral, and Nikita says Dupe reached millions in ARR in under 60 days; across his advising, about half the companies hit that kind of success and half fail outright.

He advises founders to be ruthless about onboarding but creative with platform APIs, because extraordinary product people know every API and invert it. Contact sync is an example: it can rank people who are not on the app yet but have many friends on it. Username exchange is unrealistic because getting a 50-person friend list could take 10,000 taps versus one. The goal is to one-tap users into the activation moment.

Nikita advises 35–36 companies across stages, and he starts by looking at analytics: how users distribute the app, the activation milestone, and what blocks it. He treats top-of-funnel marketing and in-product growth as one system, so ads, community imagery, in-app joins, and invite flows all align. Through intro.co/nikitabier he aims to return 10x his fee in the first 30 days by clearing table stakes and identifying two or three step-function changes; he often gets into Figma for live design sessions. He mostly works post-Series A or with venture-backed seed companies, has helped Dupe and Saturn, invests in about 10% of companies that seek advice, and invites listeners to follow him on Twitter.