If you want to create something that billions of people will use, a brand-new product like Google Photos, you have to be totally committed, and you have to realize that the path to success is very rarely a straight line, and you're almost certainly going to fail a few times or be thrown a curveball that you weren't expecting.

It doesn't look good, and you just have to keep moving forward no matter what. I'm David Lieb, and this is my backstory.

I grew up in a suburb of Dallas, a pretty normal childhood. I was, you know, into math and science, computers. My dad was an engineer and my mom was a school teacher, so very much a focus on, like, engineering, math, and science growing up.

I was one of the top students in school, and you know, I wanted to be first in the class, and so I got extremely competitive about that.

I really don't like losing. I went to Princeton, majored in electrical engineering there and computer science, and kind of just continued the traditional path of, like, get good grades, get good internships.

After Princeton, um, you know, I didn't really feel the urge to get a job, so I went to Stanford and joined the PhD program in artificial intelligence.

So I did that, met a bunch of cool people, ended up working on what would be the beginning of the DARPA Grand Challenge, which would turn into the self-driving cars that we have today.

But I was never a PhD researcher; like, I never felt at home amongst those people, and that led me to drop out of the PhD program and go get a job at Texas Instruments back home in Texas.

I remember sitting in my cube at TI and I read these two 25-year-old kids sold YouTube for a billion dollars, and that was the first moment that it became tangible to me that, like, oh, those guys don't look much different than I do.

They don't seem that much different than I do.

Could I do that sometime? Huh, maybe. If you were, you know, a smart engineer that aspired to do more than engineering, at least in my era, the next step was to go to business school, so you would go and get your MBA and then graduate and become a manager at some tech company.

So I pursued that path. I went to University of Chicago.

I started my MBA, and that was what I thought I was going to do. But then, the first week of business school, I showed up—you know, we're in a big room, you're meeting all your classmates—and the iPhone had just come out, and we were all trying to get each other's phone numbers so we could text each other to, like, meet up.

And one day, I typed in, like, 10 people's phone numbers and asked them to spell their last name, and then I called them, and they dismissed the call and they added my contact to their phone.

And I just thought, like, well, this is ridiculous. Why are we doing this? We have these supercomputers that can run apps now in our pockets; why hasn't somebody solved this?

You know, the idea that you could share contact information by taking two phones and bumping them together popped into my head in an accounting class one day.

If you want to share contact info with a bump, there's an app for that.

That became Bump. I emailed my buddy Andy to see if we could build it and we started working on it, we worked on it for maybe three or four weeks just kind of on nights and weekends as a side project.

Submitted it to the App Store for approval, the app went live at the end of March 2009 and we did no promotion. It just went on the App Store and some people found it.

I think the first day, like tens of people found it, and the second day, hundreds of people found it, and the third day, thousands of people found it.

And so, pretty quickly, it just started to take off. And we did a bunch of work then afterwards to try to get the word out.

We talked to press; I would go hit up any blogger that I could find, but it really was driven by word of mouth.

I think I learned a few things from the beginning of Bump. One is anybody can build anything.

Like, we had no expertise in doing this; we just figured it out. And I think the other thing that really resonated in hindsight is that when you're building a product for yourself, you can really trust your own intuition about what it should be and how it should work.

We weren't mobile app developers; we didn't know anything about building consumer products.

I just wanted to build a product that I enjoyed and I wanted to make it really good, and so all of the design decisions that we made just kind of fell out from my intuition.

So step one is: find a problem that bugs you about the world, and step two is: go build it.

In the early days of building a product, you have to do everything.

There's no team that's going to support you, there's no marketing team, there's no engineering team; it's Just all three of the founders were former engineers, and so we just divided things up.

Obviously, Andy was the best coder, so Andy wrote all the code. I worked on kind of the algorithm and the design, and Jake helped with design and the early promotion of the product.

And so we kind of needed an excuse for the school and for our parents to work on Bump, but have it be like more legitimate, I guess, than just messing around with a side project, so we decided to apply to YC on a whim.

So our goal is to turn Bump into a verb just like Google has become a verb today, and we're looking forward to doing that this summer.

Thanks. We got into YC, and the program was to come out to California for the summer for three months and build your startup.

We got $16,667 of funding, which in hindsight is pretty small, but it was totally sufficient to come out to the Bay Area, live on a friend's floor, and build Bump some more.

Turns out, I never went back to business school, never finished that degree.

I had been on the traditional path of going to business school, becoming a manager at a tech company, and then this small, little random moment of having an idea for an app took me on a completely different career journey.

Throughout the batch, the product kept growing, and by the time Demo Day came around in August, we were actually the number two app on the entire App Store, so we were like the most popular app in the world other than iGlowStick, which always will bother me.

Our Demo Day pitch was pretty simple: I showed a map of Bumps happening live around the world, and the map was just lighting up every second, and so everybody realized the Bump was like the hottest thing around.

MC Hammer invested, people were dressing up as Bump as their Halloween costume. That fall, Steve Jobs put our icon on a slide on one of his pitch decks.

We were the hot app of the time. The problem is, um, people didn't keep using the app, and I think this is a thing that took us many years to figure out.

The frequency of use of Bump was too low for the value that it provided people.

If you think about kind of a 2x2 grid where there's high frequency and low frequency and then high value per interaction and low value per interaction, obviously you want to be in the high-frequency, high-value box.

That's the best place to be, and it's okay if you're in the low-frequency, high-value or high-frequency, low-value, but there's one box you don't want to be in, which is the low-frequency, low-value box, and that's where we found ourselves with Bump.

And in hindsight, you know, I realized we made a ton of mistakes. We didn't ever have a credible idea for how we would make Bump a business.

We hired a bunch of people to just keep up with demand, and you know, in hindsight, we probably should have hired a lot slower.

We spent time going to conferences; all the VCs wanted to talk to us, so we took all the meetings.

We raised too much money too quickly because we could, and it prevented us from realizing the real problems that we had in our business.

All the classic mistakes that, you know, we at YC now try to teach founders to avoid, uh, we made them all.

Ultimately, we finally kind of came to terms that, oh, this might not work. Like, what are we going to do? And so we went back to the very basic YC startup advice, which is: talk to your users.

If you don't know what to do, go talk to your users.

We were many years into the Bump journey. We'd burned $10 million or more, probably probably more, probably $15 million, and I asked our team, "Give me the email addresses of the top 100 users of Bump around the world," and that day I just emailed them all personally and I said, "Hey, I'm the founder of Bump.

You're one of our best users. Would you mind talking to me on the phone today?

" And that day, I ended up talking to maybe like 20 or 30 of those people, and I learned some interesting things.

We knew from our data that they were all primarily using Bump to share photos and not share contact information, but I didn't understand, like, why?

Like, what was the context? And by talking to those people, I heard over and over and over again that day that they were using it to share photos of their family with their family members, and that was something that was a big surprise to us.

We had no idea. And when we thought about it, we realized, well, the Bump product is actually like not the best product in the world if you were going to solve that problem.

Um, to solve that problem, you don't want to like physically have to touch your hands together in order to give somebody some photos.

So we realized, well, if we're going to solve the photo sharing problem, like, we should build a really great app for that.

Uh, and that's where we kicked off building our second app, which was called Flock.

So with Flock, you would just install the app, you'd go live your life, you'd take photos as you normally do, and it would figure out which photos you took in the presence of your actual friends.

And we did this with geolocation, with social graph, a bunch of other inputs, and help you share those photos with the people that you took them with.

And so we would go talk to people and ask them, "Hey, what do you think of Flock?" Like, "Isn't it great?

" And they would say, "Oh yeah, yeah, I love it, it's so great." And then it turns out they weren't using it.

I think it's really hard to get people to tell you hard things, and so the best way to do it is to actually look at quantitative data.

Like, look at the logs, see if people are actually using it, look at your retention curves. These things are really simple, especially when you have very few users.

You know, this was a pretty tough period, right? We had all these expectations on us. We had 150 million users of Bump. It was the most popular app in the world still, but we knew the secret, which is it's going to fail.

Our second attempt, Flock, also didn't work. We're running out of money, and so I felt a lot of pressure.

This was, you know, probably one of the most challenging periods of my life, and I just wanted to figure out a way to keep this airplane from crashing.

And so we went to PG, we talked to Paul Graham, and we told him what's going on.

Here's our status. He made us kind of think a little bit bigger than we were thinking. So we showed him Flock and we said, "We got this really awesome photo-sharing app, but nobody's using it." And he pushed us.

He said, "You know what?" you should just replace the entire photos app on the phone.

Then people will probably use your sharing products. And at the time, I thought that was preposterous, like, how could we ever do that?

That would be impossible. But as we got thinking about it, we realized, you know, the photos app on the iPhone isn't that good.

And if we were designing it, we would make it way better.

So we decided to go build that product with, you know, four or five months of runway left. We called it Photo Roll.

We never launched it; it ran only on my iPhone because we never scaled it to work on anybody else's. But it was the product that we wanted as users of photos and iPhones at the time.

So I went around and tried to talk to a bunch of companies to see, you know, who might want to buy Bump, who might value this Photo Roll idea that we had.

And it really was hard for me to accept, but like, no one wanted Bump; it was not valuable.

But this idea of Photo Roll and the demo that we had built was something that several companies indicated, you know, "You might be right, like maybe that should be the photo app of the future." Ultimately, we sold the company to Google, and the plan was to go turn Photo Roll, rebuild it with Google technology, and make it Google Photos.

I was actually pretty optimistic about what we could do.

We had an aligned plan; like, it was very clear we were going to go build this product, and I was just certain in my bones that it was going to be a winner.

Then we showed up to Google on the first day and there had been a reorg and all of the plans that we had agreed upon with our sponsor were out the window, and we were tasked to go work on Google+, the social network, and I basically just believed so strongly, based on all of the time that we had spent working on Bump, that I was certain Google Photos would be a billion-user product if we built it, and I kind of just didn't do what my boss had asked me to do, and it created a ton of conflict.

I just kind of did it on the back burner. I would go every day and do my day job, and then in the afternoon I'd go spend some time with our lead designer and design Google Photos, and we did that over several weeks, and we got to a point where we had a plan for what we would go build.

I started to get some support from the engineers on the team.

They all looked at what we were working on and they said, "Oh, I want that product too. Why don't we build that instead of this social network that we don't necessarily want?

" And we just kind of slowly got some groundswell support, despite the fact that our bosses really were explicitly telling us not to do that.

I think a big part of this for me personally was we had had this enormous opportunity with Bump, and then by many measures, we'd blown it, right?

We ruined it, and I didn't want to be a loser. I didn't want to have spent the last five years of my life building this thing that ultimately wouldn't have any legacy in the world, and I saw the opportunity with Google Photos to build something that would be very long-lasting, that would make all the work that we did on Bump, the 30 people that worked at Bump over those years, make all of their contributions actually matter in the world, and I just couldn't let it go, like I couldn't say no to that opportunity.

I didn't want Bump to be a failure, and so I just didn't take no for an answer. When when the bosses told me, "You know, we're not doing Google Photos," I said, "Okay, boss," and then I went and worked on Google Photos.

So, ultimately, they told me to leave the team, right? In fact, twice I was fired from the team and told to go find a new role at Google.

I called in a bunch of favors. We had a bunch of friends throughout the Valley that knew various people at Google, and I just used every single pathway I could to get support to build Google Photos, and eventually we won the battle and we were given the green light to go build it.

So, I think this is a lesson that took me a while to learn, but you can take way more risk than you think, and if you do, there's big upside available.

So, once we got the green light, it was a team of like 20 of us from Bump latched on to maybe a team of like 100 core people from Google with access to all of this incredible technology, and we were able to build all of Google Photos basically from scratch in about nine months, and we launched it at Google I/O in 2015.

I'm thrilled to introduce a brand new product, Google Photos.

The core idea of Google Photos was a home for all of your life's Memories.

It was a place you could keep your photos and videos. We did very early pioneering work on AI to make them searchable, organize your photos by face, create things, do editing for you.

It was this idea of building a photo assistant for every human on the planet.

How David Lieb Turned a Failing Startup Into Google Photos | Backstory

Videos to help you build a successful startup. Subscribe for startup advice, founder stories, and a look inside Y Combinator.

What is Y Combinator?
We invest $500,000 in every startup and work intensively with the founders for three months. For the life of their company, founders have access to the most powerful community in the world, essential advice, later-stage funding and programs, recruiting resources, and exclusive deals. 

Visit ycombinator.com to learn more.

Video Team:
Zach Both - Video and Content Lead
Ryan Loughlin - Senior Producer



How David Lieb Turned a Failing Startup Into Google Photos | Backstory

Scripts