- The speaker discusses their experience with starting two companies, Fog Creek Software and Stack Exchange.
- They emphasize the importance of finding a cool place to work as a programmer and deciding to build their own workplace.
- They talk about the initial business model of Fog Creek, which involved consulting and building websites to generate revenue.
- The speaker mentions the dotcom crash in 2000 and how it affected the consulting market.
- They explain how they pivoted to selling bug tracking software, which gradually increased their revenue.
- The speaker highlights the slow but steady growth of Fog Creek and the launch of various products, some successful and others not.

- Stack Overflow was created as an alternative to Experts Exchange
- It initially started as a side project but grew rapidly in popularity
- The website saw continuous growth in daily visitors
- Stack Overflow surpassed 6 million users and became a popular resource for developers
- The speaker discusses the difference between "get big fast" companies and organic growth companies
- Stack Overflow is considered a "get big fast" company due to its network effects and rapid growth
- Organic growth companies focus on slowly gaining customers and breaking even early on
- The speaker emphasizes the importance of value and frugality in organic growth companies
- "Get big fast" companies have a small chance of becoming a multi-billion dollar business, while organic growth companies have a higher chance of becoming a multi-million dollar business.

- The speaker discusses the choice between bootstrapping and taking venture capital (VC) for business growth.
- They mention that there are millions of businesses making around a million dollars and only a small number of "get big fast" companies.
- The speaker shares their experience with Stack Exchange, a question and answer website, which they initially bootstrapped for two years before deciding it was a "get big fast" business.
- They explain the importance of delaying raising money as long as possible to increase the company's value and minimize dilution when raising funds.
- The speaker also mentions their success in raising money from their top choice VC, Union Square Ventures, due to their reputation and momentum.
- They discuss the growth of Stack Exchange and the need for a land grab in their industry, as well as the launch of another product called Trello.
- They explain how they use profits from their cash cow, Fog Bugs, to invest in Trello without needing external VC funding.
- The speaker emphasizes the benefits of having traction, an established team, and a proven product before seeking VC investment.

- The speaker discusses the importance of deciding whether to be a slow growth or get big fast company
- They share their traumatic experience at Juno Online Services competing against AOL
- AOL had a successful land-grab business with network effects and lock-in
- Juno had the same potential but didn't invest in marketing like AOL
- The speaker encourages the audience to decide their approach and control their own destiny by either bootstrapping or getting big fast
- They suggest using a small, reliable business as a stepping stone to build the next stage of growth.

The speaker discusses their experience starting two companies, Fog Creek Software and Stack Exchange. They highlight the importance of finding a cool workplace and building their own. The initial business model of Fog Creek involved consulting and website building, but they pivoted to selling bug tracking software. Stack Overflow was created as an alternative to Experts Exchange and experienced rapid growth. The speaker discusses the difference between "get big fast" and organic growth companies, emphasizing the value and frugality of the latter. They discuss the choice between bootstrapping and taking venture capital, sharing their experience with Stack Exchange. They emphasize the importance of traction, an established team, and a proven product before seeking VC investment. The speaker encourages the audience to decide their approach and control their own destiny, suggesting using a small, reliable business as a stepping stone for growth.

Joel Spolsky, founder of StackExchange and Fog Creek Software, at Y Combinator Startup School on October 20, 2012.

Presented in Stanford Memorial Auditorium by Y Combinator and the Stanford Technology Ventures Program.

How many of you are going to make some kind of a startup after this event when this is done?

How many of you are going to be 1 of those 15 startups every year that hits 100 million in revenue and deserves VC funding?

I made something which a lot of people don't understand is that these 2 companies that I have, Fog Creek Software and Stack Exchange, and I'm going to tell you that sort of just in chronological order to try to explain it.

And there's all kinds of interesting lessons in there, I think, for different ways to create a startup and to bootstrap a startup or take funding.

Almost every talk that I've heard today has assumed that you're going to be getting money from VCs.

And there isn't enough money in all the VCs in California to fund all your companies.

And so if you all kind of bang your heads against that door, some of them are going to get bloodied.

I'm sorry, I'm really bad with metaphors.

Michael Pryor and I started Fog Creek Software in the year 2000 after we had been working together at a company called Juno Online Services, which was a free ISP.

I was already in my mid-30s by the time I started Fog Creek.

The point was that I couldn't find a cool place to work.

I couldn't find a place where I wanted to work as a programmer.

And I finally gave up on it and decided that if I was going to have a career that I found enjoyable as a programmer, I would have to build my own workplace.

And Of course, the irony is that I don't get the program anymore.

But the very beginning of Fog Creek, we rented this office, and that was my desk.

And I literally took home everything of the company in my knapsack at work at the end of the day for work.

But I wanted to have a place to go to work.

And it was me and Michael, and we hired a couple of people very quickly.

Now, we had, the reason we knew we could start a business, even though I'm kind of a scaredy cat and not an entrepreneurial type, was that I had seen a whole bunch of other businesses in those days, So this was the first dot com boom.

And this was sort of a period, much like today, when people were starting all kinds of idiotic businesses to sell grilled cheese sandwiches and whatever crazy, you know, Instagram for squirrels, except it was called something else in those days.

And all these idiotic companies, what you could do is you could look at them and say, wow, if I did that company and I just didn't do these 2 stupid things, but I did them in a smart way, then that would definitely be a better business than this 1, right?

So for me the model was a company called ArsDigita, which was a consulting company that Phil Greenspan created which collapsed after they took VC, but everything did really.

And we were gonna make 2 tweaks on the ArsDigita model.

So he had this model that was like, we're going to do open source software and then consulting based on the open source software.

And my tweak was it was going to be proprietary software and consulting based on the proprietary software.

And then you'd have another source of revenue, was the theory.

It doesn't really matter what my theory was.

We worked on that for about 15 minutes before we effectively pivoted to something else.

But the most important part of our business model is we weren't gonna spend any money until we had some consulting revenue.

And so this was a time in the late 90s when there was absolutely unlimited demand for anybody who could put together, you know, a left, if you could type a left, a less than and a greater than sign to make HTML on the keyboard, then you could bill about $200 an hour.

And if you could actually make an HTML page, that was like double that.

And there was sort of essentially infinite demand.

There was a company called Scient, which fortunately is now completely gone.

But in those days, they hired thousands of computer science students and stuff like that to build web pages for people.

And Scient used to brag that they only accepted 10% of the clients that wanted to hire them.

So if you wanted Scient to build your website, you had to apply and then you had to get approved by the Scient board of whatever.

And we thought, you know, we'll just take some of that overflow of all that demand, build some websites for some people, make a few hundred dollars an hour, you know, with 2 or 3 of us, and build up enough money in the bank to basically bootstrap the business.

We had about 2 months of consulting, and I think we might have accumulated $250,000 in the bank or something like that, and that was really, really helpful.

So we started in September 2000 in this very office.

The consulting market, the dotcom crash occurred in November of 2000.

I know the exact month because in that month if you look at all these consulting companies, there's Scient, Razorfish, Viant, March 1st, there was a whole bunch of them.

And if you looked in November 2000, their billings went down by 90% in 1 month.

And they didn't know it, because they thought that this was just an unusual lengthening of the sales cycle, meaning it was taking longer to close new deals.

They didn't realize that this market had completely collapsed.

And they didn't really decide that the consulting market was gone and never coming back until April of 2001.

And In the meantime, they had these big armies of consultants, and they burned through every penny that they had in the bank, paying these armies of consultants to do nothing, because they had no clients.

They had no clients because everybody got scared that the dot-com crash was going to happen and stopped hiring consultants to save money.

We had accumulated a little bit of money, and we luckily hadn't hired very many people.

So, we didn't waste all that money just keeping people employed when we had no money.

So, it was just me and Michael for a long time as the co-founders.

And we moved into this building, which conveniently my grandmother owned, and I was in charge of managing it.

So I just sort of moved in and continued to pay her rent for a while.

And we continued to sort of build a business really, really slowly.

And because there was no consulting business, what we started selling was FogBugs, because the only thing we could find, which is a bug tracking software written internally in VB script.

And we started selling FogWags version 1.0.

And the first day we sold it, there's like $2,000.

Somebody bought a site license from Brazil or something.

And in fact, what we were noticing is I think in the first few months, we were making somewhere around $5,000 a month, I think, is the right number.

And you know, eventually we paid my grandmother for all the rent that we were consuming.

And as time elapsed, The neat thing was that this number was going up every single month and it was going up very slowly But it was going up and it was in a really really bad market but we were selling software and it just got better and better and better because we worked on it and Our revenue went up to when it hit $15,000.

We could afford to pay Michael a salary, and I had some savings in the bank.

So $15,000 a month is, I always just automatically calculate $10,000 per person that needs a salary, and all our overhead expenses, including the rent, was $5,000.

So we then hit 25,000, and I started taking a salary, And when we hit 35,000, we hired a third guy who came on to start answering the phones and doing some tech support and helping us with some of the coding tasks that we had.

This was I'm guessing about 2003, so like several years on.

It was a really long slog before we got our first legit office.

And We built it out really nicely, and we were always obsessed about giving programmers air on chairs and 30-inch monitors and private offices with doors that close.

And when we built this office, I think we had room.

The first time we took this office for 7 people.

And eventually, we doubled the space that we had there and we made room for I think 12 people or something like that.

We all ate lunch together at this big table.

We took summer interns every summer and that's how we got our recruiting pipeline.

And again, Fog Creek just continued to grow.

And this is the last time I remember getting the company together for an all-company picture.

But I think we're now at, let's say, 40, 45 people.

So that's 1 kind of company, and that's Fog Creek.

We continued to sort of try to launch new products every single year.

We had this idea that bug tracking was not really the be all and end all of products.

We had a bunch of summer interns launch a remote desktop product called Copilot.

That actually more than paid for itself and now earns us sort of a nice dividend.

FogBugz was obviously the big breadwinner and is now a very large product.

Some of it worked, some of it didn't work.

We launched a job board for the Indian programmers market and made 50 rupees total ever, selling 1 job listing.

I wanted to kill Experts Exchange, which you all probably know about, right?

Anyone heard of that Stack Overflow thing?

So the goal there was to kill experts exchange because they were evil.

And what was cool about Stack Overflow is you could explain it that simply and everybody understood what you were talking about.

You didn't have to go into any detail of how it would work.

And I couldn't find anybody internally that wanted to work on that, so I got together with Jeff Atwood, who's another blogger known as Coding Horror, and said, you know, will you join us?

This'll be sort of a 50-50 Fog Creek, Jeff Atwood joint production, to do Stack Overflow.

And I kind of thought of this as sort of a side bet, like let's get Jeff working on this.

If it doesn't take off, doesn't really matter, but hopefully I can get rid of this experts exchange problem.

But Stack Overflow did really well, as it turns out.

Unlike the other speakers, I am perfectly happy to number the numbers on my y-axis.

And I'll even tell you, the growth here was insane.

Like, literally every single week, we had more visitors than the week before.

At the beginning, I think it was the numbers were crazy.

We started out with like 30,000 daily visitors.

And that's about the same number of people as I would get on a Jolan software blog post.

So I knew that that was the Jolan software audience coming initially.

And that took 10 years to build that audience, by the way, writing a blog.

And You would look at the Google Analytics every single week and it would just get a little bit higher and of course it went down on the weekends and then it would just go a little bit higher and at some point it was growing by more than a jaw on software like every week and then it was growing by a jaw on software every day and It just grew faster and faster and eventually encompassed the universe.

Now, at 1 point, when we were around 6000000 on here, so this measure is global unique monthly visitors.

That measures people that just visit your site once a month and we place a cookie and all we do is count the cookies on Google Analytics.

So there's maybe 2 users on 2 different, maybe 1 person on 2 computers.

The other thing about Google Analytics is that the cookie is placed on your domain and we have 4 domains.

And we may be double counting people that visit 2 of our sites if they go to 2 different domains.

And when we hit about 6 million users, we had a meeting with the folks in developer relations at Microsoft and we said, hey, how many developers do you think there are in the world?

And we're like, okay, we got two-thirds of the developers in the world.

And then we hit 9000000 and then we hit 10000000 and then we hit 20 million.

Now we're at about 25 million of the world's 9000000 software developers.

So there's 2 kinds of businesses here that are completely different, actually.

There is 2 ways you can build a business.

And everybody up until now today, from Zuck onward, has been talking about the get big fast business.

And Stack Exchange is indeed a get big fast business.

Whoa, that's not going to work anymore, is it?

The stackexchange.com domain is now growing at 350% year over year, which is, I don't know, not the fastest thing in the world, but it's pretty frigging steady as you saw from that chart.

I wrote a blog post on Jolent Software a long time ago, like in the early 2000s when I didn't have anything else to do because the economy was collapsing.

It was called Strategy letter 1 because I was very self-important and pompous then, unlike today, of course.

And I was writing about how there's 2 kinds of companies.

There's get-big-fast companies like Amazon, and then there's organic growth companies.

And I had just read the story of Ben and Jerry's, So they were an example of an ice cream company and I kind of compared and contrasted these 2 different ways of growing company And what's interesting of course is that stack exchange and fog Creek have you know?

Kind of exactly the same characteristics stack exchange is a get big fast company and Fog Creek Software is an organic growth company.

So first of all, how do you decide which kind of company you should be?

Like how many of you want to get big fast?

You have to look at the actual market that you're in.

A get big fast business is when you're in there for, there is some kind of a land grab.

There is an attempt to capture, uncapture territory before anybody else does the same thing.

Ben and Jerry's was not, because Ben and Jerry's was just ice cream, You can't capture the entire ice cream market.

There already isn't ice cream market, right?

This is like you're trying to new technology trying to land grab virgin territory.

In an organic growth business, you already have a million competitors and all you're doing is trying to claw them away 1 at a time patiently from your competitors.

The reason you try to get big fast and you go for the land grab is because there are network effects.

So you all know what network effects are?

The idea of network effects is the more users you have, the more valuable your network is.

And so a network of, let's say if Facebook had 10,000 users, that's not very valuable.

And in fact, it would be extremely difficult.

Facebook was a classic land grab because nobody's going to be on there if they can't communicate with their friends.

The same reason I can't get anybody to use Kik, because there's nobody else on there.

I mean, there's like 10 million people on there, but it doesn't really help.

And So a network effect says basically the value is increasing as you get more and more people on there because the value is a function of n squared, right?

The other thing about get big fast is that network effect creates lock-in, right?

Nobody's gonna leave Facebook for something that's 20% better or 50% better or even 100% better because their grandma is still looking for the pictures of the puppy.

And she likes the pictures of the puppy that you upload and if you stop putting them on Facebook, she's gonna call you.

It doesn't matter if the other thing is more technologically innovative and cool.

Now, in an organic growth company, you don't have these things.

And so you don't really care about that and you don't need to go for a land grab.

And a land grab won't help you because you're not grabbing virgin territory.

You need to be really big in order to create real value.

And with organic growth, you're in there for the long haul.

It took us 10 years to build FogBugs into a respectable business.

And The reason is that there are a million bug trackers out there.

With organic growth business, unless you have infinite money because you were born rich or something, you need to break even right away.

You need to find a way to break even right away.

We did that by consulting, by essentially sort of whoring ourselves out for a couple months to raise a little bit of cash.

And you wanna start getting money from clients on day 1, and you wanna start kind of obsessing about, I need to make this business survive so I can be there to slowly pry away customers, 1 dead customer at a time from my, and 1 thing about an organic growth business is you don't go in here unless you have a product that's valuable even with 1 customer.

So the ideal business is something where the first person that downloads it and starts using it is already getting value out of this thing.

You run your business completely differently.

So when you're trying to get big fast, if you have a problem that can be solved with money, money is cheap, which I know sounds weird, but in this environment, if you have a successful get big fast business, It's very easy to raise money.

There's lots of capitalists that are looking for places to put their money.

And so you use that money to try to solve problems, even if it's a very expensive way of solving them.

But in an organic growth company, you're trying to survive.

You're just trying to stay around for as long as you possibly can.

So you try to be frugal and cheap about everything that you do.

Similarly, when you're trying to get big fast, you put up little signs all over the office that say move fast and break stuff.

And the idea is to make lots of mistakes and it doesn't matter, you gotta be moving fast, you can't be scared in any way, you just gotta be running down the hill.

With an organic growth business, those mistakes can kill you.

And when they kill you, you don't have any way to come back.

The thing that scared us at Fog Creek all the time is that if we ran out of money in the bank, we would have to close.

We did not have a source of a million dollars to stick into the business to survive 1 more month.

So you're much more careful in an organic growth business because you have to be because this is your only business.

And that means that the get big fast business, you have a tiny, teeny weeny, very, very, very small chance of it becoming a $10 billion business.

And I just made up the number 10 billion, but really, really big business.

And if you consider that of all the Y Combinator, let's say you all apply to Y Combinator and you all get in.

You know, 2 of you, actually of all the people in this room that apply to Y Combinator and the 60, 70, 80, I don't know how many get in, you know, maybe 2 or 3 of them have a chance of being a $10 billion business.

And that's being really, really optimistic.

So your chances of you, of making a $10 billion business this way by starting out trying to get big fast are vanishingly small, I'm afraid.

And some people like those odds, so good for them.

Some people, incidentally, all the people that make $10 billion couldn't care less.

They're doing it because they're scratching an itch.

They're trying to solve some problem that they have and they're monomaniacally focused on that.

It's never because they want to make $10 billion.

Now, with an organic growth company, though, if you're just reasonably smart and you pay attention and you never make a terrible mistake, then you're going to make a nice $10 million business.

And the number means $10 million means $10 million in revenue a year.

And at some point, that's just going to be a million to operate it and 9000000 that can go in your pocket if you own it.

And you have a really good chance of being able to build that in 5 to 10 years in organic growth business.

When I used to write a column for Inc Magazine, and I always wanted to write something about VCs, and my editor always said, nobody, no entrepreneur takes money from VCs.

It's such a narrow world of people that get investments from VCs.

It's such a tiny sliver of entrepreneurship that if we write about that in our pages of our magazine, we get hate mail because Nobody cares.

There are millions of businesses out there that are making somebody around a million dollars.

Are you bootstrapping or are you taking VC?

Again, there are millions of organic growth companies and there are a very, very small number of get big fast companies.

I wrote a few hundred but that's like the entire universe of companies in the world.

There's like 1 every decade or so from Silicon Valley.

And not being able to decide is what's really gonna kill you.

So let me see if I can sort of finish a little bit of the story about Fog Creek and Stack Exchange.

See all Y Combinator transcripts on Youtube

Joel Spolsky at Startup School 2012