In 2009, David Lieb put two phones together and they swapped contact information. Not via Bluetooth pairing. Not via QR codes. You just bumped them. The physical gesture - almost absurdly simple - turned into the most downloaded app in the App Store's first billion downloads. Bump was everywhere. On 20 to 40 percent of all smartphones worldwide at its peak. TIME magazine called it one of the apps that helped define the iPhone. Apple put it in commercials.
None of that solved the actual problem: Bump never made money. After raising $19.9 million from Andreessen Horowitz and Sequoia, after 150 million users and the kind of press coverage that most founders would trade years for, the company could not find a repeatable path to revenue. In September 2013, Google acquired the entire 25-person team for around $35 million. The app was shut down by December.
What happened next is the part most people don't know. The Bump team had been quietly building a photo-sharing app on the side. Google looked at it, recognized something, and gave the team room to develop it further. That side project became Google Photos.
"Trust your gut on where this thing should go, and allow it to go those ways. During the journey of Bump, there was so much pressure on us because we raised a lot of money from fancy investors that it probably made it harder for us to see where this product wanted to go - and ultimately it wanted to become Google Photos."
- David Lieb, Y Combinator Backstory Podcast, December 2024Lieb led Google Photos as Senior Director for nine years. He built and launched it in under nine months with a small acquired team. By 2019 - less than four years after launch - it had crossed 1 billion users, one of the fastest growth rates ever recorded for a consumer product. The AI search, the face recognition, the automated memories you didn't ask for but can't imagine living without: Lieb's fingerprints are on all of it.
In 2022, he left Google to join Y Combinator as a Visiting Group Partner. By 2024, he was a full General Partner. The kid who went through YC as a founder in the S09 batch is now one of the partners deciding which founders get in. The companies he's advised are collectively worth over $10 billion. He has conducted more than 500 office hours across YC batches.
Princeton Pitcher. Stanford AI Researcher. Accidental Entrepreneur.
Lieb grew up outside Dallas, Texas, the son of an engineer father and a teacher mother. He pitched baseball at Princeton while earning a degree in Electrical Engineering - a combination that suggests either unusual range or a very specific kind of stubbornness. After Princeton, he went to Stanford's AI Lab, working with Sebastian Thrun, the researcher who later founded Udacity and led Google's self-driving car project. The circles in Silicon Valley trace back to people like Thrun; working with him at that moment was not incidental.
He started an MBA at the University of Chicago Booth School of Business. Left to build Bump. That decision - dropping a prestigious business degree to join a startup accelerator - turned out to be one of the more consequential ones he'd make. YC S09 included Bump. It also included the people and network that would shape the next fifteen years of his career.
The same institution that funded Lieb as a 27-year-old founder now writes checks based on his recommendations. He is, in the most literal sense, a product of Y Combinator who became part of the machine itself.
What Failure Looks Like From the Inside
The post-mortems on Bump focus on the monetization problem. But Lieb's own telling emphasizes something harder to see from the outside: the distorting weight of too much money. When you've raised from Andreessen and Sequoia, when you have 150 million users, when Apple is putting you in commercials, the pressure to find a business model in the existing product becomes enormous. It crowds out the signal from what the product is actually telling you.
What Bump was telling them, it turns out, was that people wanted to share photos. The contact exchange was the original use case, but the behavior that stuck - the thing people came back for - was photo sharing. They built a separate app called Flock around that behavior. Google saw both apps, bought the team, and shut both products down almost immediately. The insight, the team, and the technical foundation all survived. They became Google Photos.
The Problem with Success
150M users but no revenue model. Fancy investors creating pressure to monetize within the existing frame. The product was pointing somewhere else the whole time.
The Hidden Signal
Users kept coming back for photo sharing, not contact exchange. The side project - Flock - was the real product. Nobody saw it clearly until after the acquisition.
"The product wanted to become Google Photos."
Nine months to launch. Four years to a billion users. Lieb now opens every YC batch by telling this story to founders who think they understand where their product is going.
The lesson Lieb now teaches is precise. Cohort retention is the signal to watch. If you retain a percentage of each new cohort over time - a flat line rather than a declining one - you have product-market fit. You can then make the curve go up. If the line slopes down, no amount of user acquisition fixes it. The problem is in the product.
His other rule: don't raise too much money. Not because money is bad, but because money is a drug that makes it harder to feel what your product is telling you. You spend it. You hire more people than you need. You optimize the existing direction instead of finding the right one. Bump had $19.9 million and 25 people and couldn't pivot fast enough. That's the lesson he carries into every office hours session.
Nine Months. Nine Years. One Billion People.
Building Google Photos inside Google with a team that had just come through an acquisition should have been impossible. The incentives at large companies don't favor small teams starting from scratch. Lieb made it work anyway - launching in nine months and then managing the product for nine years. The philosophy he brought from Bump translated: deep focus on cohort retention, obsessive attention to what users were actually doing rather than what they said they wanted, and a willingness to use AI not as a buzzword but as a specific tool for specific problems.
He defined AI's role in a product with a question: does the AI uniquely make this better? Not does it seem impressive. Not does it check a box. Does it solve something that couldn't be solved otherwise, or scale something that humans couldn't scale alone. Google Photos' face recognition, its ability to search for "photos from the beach in 2019" and find them accurately, its automated memories and highlight reels - these were answers to that question. AI did something that a human-curated album library couldn't do at 1 billion users.
By 2019, Google Photos had a billion users. By the time Lieb left in 2022, he had built something that nearly everyone with an Android phone - and a substantial fraction of iPhone users - used to manage the most personal archive of their lives.
"I think Google's lawyers don't understand how much they are kneecapping Gemini's chances of beating ChatGPT. Personal context and habit is how you win here. Google is perfectly positioned with Gmail and Google Photos data - but they are blowing it."
- David Lieb, @dflieb on XThe Mentor Who's Done the Thing
Most venture investors have either built a product or invested in products. Lieb has done both at scale. He built Bump to 150 million users, scaled Google Photos to 1 billion, and has now deployed capital into companies like Flexport - his best-performing investment by his own account - while sitting in office hours with founders who are figuring out which problem to solve next.
At YC, Lieb focuses on early-stage founders in consumer technology, mobile, and AI. His advice is consistent across all of them: hire less. Raise less. Watch your cohort retention curve. Follow where the product is pointing, even when that direction feels counterintuitive given the money you've already raised and the narrative you've already told.
He opens each new batch at YC by telling the Bump story in full - the iPhone commercials, the 150 million users, the monetization failure, the acquisition, the shutdown, the side project, the billion-user outcome. It's not a triumphalist story. It's a story about how hard it is to see clearly when you're inside the thing, and how the product often knows more than the founder does about where it needs to go.
Don't raise too much money, because you'll just spend it. When you have a lot of money, don't waste it on stupid things. Don't hire too many people. Stay focused.
- On startup discipline
You'll know you understand the customer problem enough when you can predict 75% of what a customer tells you.
- On product-market fit
Does the AI uniquely make this better? That's the question every product team should be asking before shipping AI features.
- On AI product strategy