They do have wonderful things to tell us about their experience, and I won't get in the way and let them start. So welcome back to Stanford. Thank you so much.

Thank you, Tina. Thanks so much for having us, Tina, and Stanford. It's great to be back here. I remember... how many years ago now, four or five... sitting in this exact room and watching people stand up and give advice about entrepreneurship and their experience, and it's a little surreal to be standing up and giving back. But it's a really exciting opportunity because, I think in the past year or so since we started what would become Instagram, we've learned a lot. And today what we want to do is go through a series of myths that we think we had learned along the way, or thought were true along the way. And as we did Instagram and as we went through the process of founding this company, we learned that not all of them were necessarily true. So the big caveat here today is, although we're saying all this stuff, experience is what matters. And going through your own experience in a startup is really what helps you debunk these myths as well. So this is our chance to share some learning with you guys. My background, obviously, I went to Stanford. Mike went to Stanford. I studied MS&E, Mike, you studied... SymSys. ... SymSys here. And that was the beginning of our entrepreneurship experience in the Mayfield Fellows program. Like Tina said, we both had really amazing internships then that got us to get interested in entrepreneurship and get excited about doing it when we got out. And both of us after a year or so of working in a larger company decided we wanted to do something. And hopefully today, through that experience, we can shed a little light on what we learned. What we're doing today, Instagram, is kind of interesting because it came out of something we were doing before that didn't quite work. How many people here have actually heard of Instagram/use it? OK. Awesome. Most of the room. How many people have heard of Burbn/use it? Used it. Yeah, like three people. That's awesome. So that's why we started working on Instagram because that's basically the number of hands that went up in the room when we were working on it. Burbn was this check-in app that lets you check into different places, and while you were doing that allowed you to share pictures or videos of what you were doing. Long story short, we worked on that for a little while and then realized it wasn't really going anywhere. But the thing people loved the most about it was actually sharing images of what they were doing. So today, Instagram has about a little less than four million users all sharing images of what they're doing out in the real world through their iPhones on a daily basis. How many mobile photos do we upload per day about now?

It's like... Jesus, six a second, more or less. So whatever that times... 

Yeah, it's a lot. And this is pretty awesome to be in this position only six or seven months after having launched. But the myths we're going to talk about today, I think, really helped us get to the next level and start Instagram by learning that those myths weren't necessarily true. So to start, I think Mike's going to bring you through the first myth.

 you're dealing with the bucket of uncertainty that is being an entrepreneur and getting started, you want to latch on to things that you've seen before. We really quickly learned that you just cannot really learn to be an entrepreneur from a book, a blog or a talk, and it turns out that a day on the job was worth a year of experience and what happens is the collection of experiences and knowledge you can get from those sources are super important. And I'm not dismissing them entirely as something that you should just ignore, but that first day when you're starting to make those decisions where the data isn't really there and there hasn't been a blog post posted to Hacker News that was like, 'Deciding what to do on the first day of startup' or 'Making this really tough decision', it turns out a lot of it is very specific to your situation and all you can really learn to do beforehand is try to deal with that uncertainty. So making snap decisions or quick decisions in the face of a lot of uncertainty. We'll hit up on situations early on where we weren't sure if we were going to take Instagram a follow model, for example like Twitter, or more friendship like Facebook. And there is just no blog, book or talk that we could've ever really seen beforehand that would have taught us to do either of those things, since that was about sitting down and saying, 'Well, what do we know beforehand? What does our gut tell us?' And trusting your gut, I think, is a theme of this talk. And so developing a better gut is the work you can invest in beforehand, and then saying, 'All right, let's invest in this. Let's stay the course for a while and really see it through,' rather than wavering for months at a time being, 'Oh, why don't we build both? Then we'll switch off,' maybe make it a preference like 'worst mistake ever,' give up on making that decision and instead make it a preference, and so and so forth where you're having these micro decisions that in the end sum up to what becomes your product basically. And we really rapidly found that, as tempting as it is to go search off for prior accounts of something similar, that's not decision is what makes a difference. But what you can be doing is doing quick projects, side projects during school, even when you're outside when you're doing a job. And most of what we learned and applied into our startup were things that we were doing on the weekends which, depending on the companies, either something encouraged or discouraged. But usually if you're excited enough about something you will find the time to work in it. The other thing is, once you do start a startup, it's super tempting to get caught up in the meta part of doing a startup, so going to entrepreneurship events and being, 'Yes, I want to talk about being an entrepreneur.' We were incubated at Dogpatch Labs, which was a great experience. We were surrounded by 30 startups, a rotating cast. We were there for probably longer than anybody else. Too long. Too long. We saw three or four different classes of startups go through that. And the successful ones were the ones that were in it 9 am and left at 10 or 11 pm and were just putting in the work, and not the ones that showed up at 10, hung around, left at 6, who in my opinion were doing a startup as a lifestyle choice because they didn't want a boss. That's not really a good enough reason to do a startup. It should be that you wake up and you're obsessed with this idea and you want to make it happen, and you're not there to hang around in this club or have this fun chat with people. And that distinction wasn't that apparent to me Day 1 because everybody's doing a startup, this should be a thing. And then one month in people were like, "You guys work really hard." We kept hearing that comment at Dogpatch, and we were. We were working the hours that we felt we wanted to throw into the startup. And I guess it's a gut check if you're finding yourself getting drawn into the meta part of the startup of being an entrepreneur, of being really excited about... Somebody said to us earlier, the phrase was like, "You can't call yourself an entrepreneur. Somebody has to call you an entrepreneur in a way." And it's true. It's very tempting to get caught up in that. And I would encourage you to step back a little bit and find out the only thing that ships products and the only thing ultimately end users care about is the product you deliver to them, not how will they talk about you in TechCrunch or exactly who your investors were or which events you attended.

Myth 2: Startups can only be started by computer science students

Reality:

- Early Twitter team never went to college, Kevin & I didn't major in CS

- Sink or swim school of engineering

- Generalists are perfect for startups

- Find cofounders that complement you

Mike: Another myth that we encountered as we started our company... we talked to our friends who were holding back from starting companies... is that startups only come from Computer Science students. Neither Kevin nor I studied Computer Science. And that's something that we're actively proud of, not because Computer Science is a bad degree by any means but because it means you can get the technical chops you need to get things off the ground to get things prototyped and shipped. We built all of the initial version of Instagram ourselves from things we mostly just were self-taught in. The early Twitter employees, none of them even went to college, and our first engineering hire didn't go to college, either. I think with Twitter, maybe they didn't finish college. Maybe they went to it. But it turns out there's things you can do in school that I think are valuable, and when you're trying to pick courses and figure out where to focus your time, the classes I look back to now and think, 'Wow, these are the ones that helped me deal with that uncertainty day to day' are the ones where Day 1 of the quarter they tell you, 'We don't know what you're going to be doing for the rest of the quarter. You'll get this at the d.school a lot and all of the entrepreneurship classes. It's your job to ask the question, figure out the question that you're going to tackle, and then answer it for the rest of the quarter.' And that's just a very different experience from, well, 'These are the 10 problems that you're going to tackle and then we'll deliver them at the end.' And, of course, going through those motions is really important as well, but having that ability to ask the question... and Kevin will talk a little bit more about this in the next one... but also just work through the rest of that quarter. And the rest of it is, the engineering we end up doing we call 'Sink or Swim School of Engineering'. So we launched on this little machine server in Los Angeles. We had no idea what we were doing. We were like, 'Well, maybe some people would sign up.' Within 24 hours, we had so much demand on that one machine that all of a sudden we had to scale out to what we now have, millions of users. None of us had touched Amazon's cloud platform at all before launching; we'd kind of heard of it but shied away from it. And it turns out that there's no motivation stronger than a bunch of people knocking at your door saying, "I want to use your product. Fix your thing." We'd put in a lot of... I don't really remember the first two months of our startup because we didn't sleep, and I think short-term memory goes out of the way. I'm told we put in a lot of late nights that were all about saying, 'What do we need to do to get our products to a place where people can keep using it, get excited about it, scale to the challenge?' And you'll learn those things because you're bright and intelligent, you started a company because you trust yourself. So having that faith and not shying away from a big challenge because you're, 'Well, what if we're successful? We won't know how to scale...' I barely really knew how to use a lot of the Linux Sysadmin stuff and now we know it really well, and if we did it again we'd have a totally different approach. But it's a little bit of zen beginner's mind: you focus on the simple, important stuff first if you're not worried about scaling ahead of time. It's really good to have friends that are Computer Science students. Absolutely. It's all the building that network. Week 1... I had worked at Meebo beforehand and I was doing mostly frontend development, so I wasn't doing a lot of hardcore scaling stuff... and I remember 8 am in the morning I'd be waking up my friends who led more normal jobs and be like, 'I have no idea what this means. How do I do this?' They'd come in, we'd buy them beer. And you build that network and they'll help you out because they're excited about what you're doing, and it becomes less about feeling like you're the entire source of knowledge for yourself.

Kevin: Right. And I think what I'd add to the original point of going to events or talks is that it turns out what you get from those things aren't necessarily the takeaway is that we're going to put up here on the board, but it's the people sitting next to you, it's the people you meet before the event, after the event. The people that you're sitting next to chatting with them about the stuff that you're doing, they'll end up being the most valuable part of your entrepreneurship experience going down the line. The fact that I remember being at a party... I think it was maybe sophomore year in college when Facebook had just moved out, and I ran into Adam D'Angelo who was the CTO at that time at this little party. And we kind of kept in touch since. And on the day we went down that first day we launched, we had all these problems. I was like, 'Who's the smartest person I know who I can call up?' And Adam spent 30 minutes on the phone with us just walking us through the basic things we needed to do to get back up. Those little events are the things that matter. So as much as you're paying attention to the stuff on the slides, make sure you spend some time after the talk getting to know the people around you.

Mike: Absolutely.

Myth #3: Finding the solution to the problem is the hardest part.

Reality:

- Finding the problem to solve is the hardest part.

- It's easy to build solutions to problems no one has.

- How do you know if you're solving the right problems?

- It's ok to solve simple problems.

Kevin: I think Myth Number 3 that I'd like to talk about, this is something I had no clue about. It was that finding the solution to the problem is the hardest part. I always thought like, you're faced with these problems that people have, you assume that you know exactly what you're going to tackle, and the hard part is finding that algorithm. The hard part is scaling that solution. It turns out... thank you, Mike... that the hard part is actually finding the problem to solve. Solutions actually come pretty easily for the majority of problems. Not for every problem, but for the majority of problems. And in our case, what we did is when we sat down and we were deciding to work on Instagram, what we did was we wrote down the top five problems people have with mobile photos because we wanted to build a product that solved problems. We didn't want to just build a cool app to look for a problem that people had. We wanted to do it the other way around. So what we did was we listed out these five problems. And I remember the top three that we circled. Number 1 was that mobile photos don't look so great. We've all had that experience... you're seeing the sunset, you take a snapshot, and it looks washed out, you can barely see the sun, etcetera. And we were like, 'That's the major problem we want to solve.' Number 2 was that uploads on mobile phones take a really long time. So we were like, 'What could we do around that?' Well, maybe if we started the upload way before you're done even editing the photo's caption, and what if we sized down the photo just to fit perfectly on the screen but nothing else? And that's the small little problem and solution that it turns out really delights people because they press 'done' entering their caption, it's already been uploaded. The third problem was that we really wanted to allow you to share out to multiple services at once. We felt like, should you have to make the decision of taking a photo with a Facebook app, the Twitter app, so on and so on, or should you just take it in one place and distribute it to many places at once? Those top three problems allowed us to really hone in on what solution we wanted to build. And that's really what Instagram became. I also wanted to say that once you have those top problems that you want to solve, you need to verify that they're actually the ones that people have. And really the way to do that is get your product in front of people very quickly and test that hypothesis. I think too many people wait a long time... and I'm going to talk about this a little later. Too many people wait a long time to see whether or not what they're working on is actually the problem people are having. And the last point is that, really, you should not be afraid to have simple solutions to simple problems. Like I said early on, I think too many people believe you have to solve things in a really complicated way. And at the end of the day, if you delight people even a little bit with a simple solution, it turns out it goes very far. That first day when we had something like 20,000 new users, I was like, 'Clearly there was a need for this that hadn't been done before,' and I'm so glad we tackled those simple problems. There's something about... in the tech community, you always want to feel like you're working on the hardest problem in the world. It turns out that a simple problem becomes very hard at scale. And that's what's really exciting.

Mike: In a way, we often... in our entrepreneurship classes we hear about the Big Hairy Audacious Goal, like, 'What's the huge chunk you're trying to bite off and tackle?' And one thing that really struck me was that that Big Hairy Audacious Goal could be bringing that simple solution to something delightful to the masses.

Kevin: Yeah.

Mike: And that, in itself, is a huge challenge.

Kevin: Yeah. And it's something we deal with on a daily basis. Mike has to wake up at 4 am everyday to reset servers and stuff. I wake up with him, but I don't actually do anything. I just say, "OK, Mike, I'm here with you."

Myth #4: Work for months building a robust product in secrecy, then launch to the world.

Reality:

- Make your product public quickly

- Build the minimum viable product that answers, "Are we building the right thing?"

- Fail early and often, make failing as low cost as possible

Kevin: Myth Number 4... This is what I was talking about before... working for months to build a robust product in secrecy, and then launch to the world. How many people have heard of stealth startup, started a stealth startup, feel like they would go into entrepreneurship and keep their ideas to themselves? I'm going to assume that everyone's going to raise their hand, because we've all heard of the stealth startup, how cool it sounds. The problem with stealth startups is that you don't get the feedback you need quickly enough. In order to test whether you're working on the right thing or not, you need to put it in front of people. And I think there are certain verticals... I'm really talking about consumer internet here specifically... I think there are certain verticals outside of consumer internet where stealth might make sense if you're doing pharmaceutical or something. But for us, getting it in front of users was the most eye-opening experience. I remember putting Burbn in front of people we didn't know and they were just, 'What is this thing? What are you doing?' We would be in a busy bar and trying to explain to them on our mobile phones and they just wouldn't get it. And that happened enough in front of people outside of our friend group that it was really clear we had to work on something different. Or at least refine the idea. And I think that's something to keep in mind as you're going about starting a startup.

Building the minimum viable product is super useful. Don't build past what you need to build to answer the questions. I think Eric Ries talks about this all the time. He says he was at his job and he was building this 3D chat client... the idea was you could basically link your Yahoo Messenger, your MSN Messenger, all these different messengers to this one 3D client... and he spent eight weeks doing it. And then they launched it. And no one used it. And he said to himself, 'Couldn't I have just done one of those platforms to prove that no one used it?' So ask yourself, how much work do you need to do to actually prove whether or not this thing's going to sink or swim? The 'Sink or Swim School of Engineering'. We should just write this book, right? It's really true, though. Everything we do at Instagram, we start by saying, what's the V1 of this feature? What's the V1 of this thing that we're going to put out in two days and test whether or not it's going to work? I think that's super important to remember.

The last point here is failing early and failing often. It's totally OK to fail in an organization. You need to fail in order to find the right solution. Often, the first thing you start off with is not the thing you end up with. And you should assume from the start, your first idea is not going to be your last. And your job is to fail your way to success. And I think what's we did in Burbn pretty well. We failed all the time. And finally we woke up one morning and we were like, 'We're failing too much. We need to move a little bit to the right.' And even in Instagram, we failed a bunch with different features and things. But the idea is you're constantly refining this original idea. It's not this 'wake up one morning and have the brilliant idea and go implement it'. You're constantly iterating on it.Myth #5: Start a bidding war among VCs with a slick pitch deck.