Breaking News: Why Didn’t Stack Exchange Work?

Stack Overflow, the question and answer site for programmers, went from 0 to 6.5 million monthly unique visitors within 2 years. So the company launched a few related sites, which did well too. That’s when they decided to create Stack Exchange, a hosted question and answer platform that anyone (who has $125 per month) can run, about any topic they choose.

As you’ll hear in this interview with Joel Spolsky, the company’s co-founder, it didn’t work. Only 20-40 of the sites created had significant traffic. Many were ghost towns. In this interview, you’ll hear why Stack Exchange didn’t work. You’ll also hear why Stack Overflow raised money and the new direction they’re pursuing.

Joel Spolsky

Joel Spolsky

Stack Overflow

Joel Spolsky  is the co-founder of Stack Overflow, a question and answer community. He’s also the co-founder of Fog Creek Software, the  bootstrapped New York City-based software company. And, until recently, he wrote an influential blog about software development, called Joel on Software.



Full Interview Transcript

This interview is sponsored by Grasshopper, The virtual phone system that entrepeneuers love, because you can use your own phones and manage it on the web. Check out It’s also sponsored by Wufoo where you can go right now to get embeddable forms and surveys that you can add to your website for free, go to And it’s sponsored by shopify. When you go to you can create a store within minutes and have all the support and features you need to make that store grow, check out

Here’s the program.

Hi, everyone its Andrew Warner founder of Home of the ambitious up start. I got a guy you are all familiar with Joel Sposkey , he is the co-founder of fondcreek and the co-founder of Stacks Exchange. Stacks Exchange is what we are going to be talking about today. It’s a question and answer platform. I built Mixergy’s question and answers site on the platform that Joel co-founded, Stacks Exchange, as have a couple of other people who have been here on Mixergy.

Andrew: Joel do you have a bit of news, what’s going on?

Interviewee: Well, I guess you’ve go the scoop here because I’ve noticed that its not quite live on the Stack Exchange blog yet. But, basically we are making….. Or may be I should give the people some background.

Andrew: Ok, Lets do that.

Interviewee: I am sure there are somepeople who never even heard of us. The goal get Stack overflow, I don’t know how to say this politely, is to wipe out Experts Exchange. I don’t know if you have ever seen that site. The Experts Exchange is a site where programmers do a community qna kind of thing. And the owners of that site for what ever reason slapped the pay wall in front of it. So programmers for years have been typing questions into google, getting told that there was a answer at Experts Exchange, and then being asked to sign up for a twenty dollar a month membership. Or what ever it was. I just thought the programmers could do alot better than that and that we could build something for ourselves that wasn’t behind a pay wall, for example. So together with Jack Atwood, we built a site called “Stack Over Flow” it went live, I quess eighteen or nineteen months ago, in September of 2008 that that site went live and we have grown in less than two years from zero to about six and half million monthly unique vistiors according to Quantcast. So pretty much huge grownth. It’s been along time since I have met a programmer that doesn’t use Stack Over Flow either as a active user and answering questions and sort of passively searching for things on google and finding the results on Stack Over Flow.

Andrew: Sorry, We’ve got a little bit of a lack. As long as we are telling the story before we get to the news, Ill be spend a little bit of time asking you why Joel.But I also want to ask you something that I have always wondered. What is the problem with, Expert Exchange, so they want to make some money, they want to charge people. Why is that a problem that you’ve made this your big mission?

Interviewee: I’s clear they have the only programmer content site throughout the entire internet that you have to pay for. Normally programmers are pretty generous with their time and with helping each other answer questions. It very unusal to find a site where there’s a private company that’s bascially trying to make money off of programmers answering each other’s questions. Not make money sort of incidentally or putting a badge or something like that, but accually make money by charging for access to questions which were provided by the community themselves. So, I think that this is basically like any pay wall on the internet. It’s basically a blind mission on the internet. The internet routes around.

Andrew: You charge for your software, why is it ok to charge for software but not charge for access to a place where people can change ideals.

Interviewee: There’s nothing wrong with the recent Exchange charging for it. I just think they will always be the competitive disadvantage agaisnt the site that makes it but great. I don’t think there’s a ethical problem with their charging execpt that the fact that a lot of their users contributed that information and contributed the answers to the site under terms that they thought they were sort of helping out the community and later became clear that they were just building a database.

Andrew: I always wondered about that. My feeling was that it was good marketing. It’s one of the reasons why I discovered Stack Over Flow. Fair

Interviewee. Was it good marketing

Andrew: No that you making it into more than just I got a new site that’s going to be free. But, It’s us verse them or our way verse their pay way.

minute 5 to minute 10

Andrew: …it’s “us” or “our way” versus their “pay” way. Help me to get my attention.

Interviewee: There’s so many programmers that have done their search for their programming question on the Internet. They found the number one results, purporting to be an answer of Expert Exchange, click through that link and found out that they need to get out their credit card and sign up for a monthly subscription. It became extremely easy for us to market Stack Overflow by just describing it as the free site where programmers can get answers to their questions as opposed to that other site where they don’t because there’s a pay wall in front of it. So, it certainly helped us explain our message.

Andrew: Here’s something that I learned from you and I think it was on Gregory Gellen’s(?) podcast. I didn’t know that when I ended up on one of their pages looking for an answer. I couldn’t find the answer, but instead, I found a link to the pay section. I didn’t know that if I just scrolled to the bottom, I could get all the answers on their site, until you say.

Interviewee: You can only do that if you come from Google. So, they’re actually showing Google a different page. If they refer, Google does show you that page. That’s just a part of their general sort of lack of concern over the quality of the Internet and just generally kind of want to trick people into paying for something but they still want to be indexed by Google. Google takes a pretty deep stance that if it’s behind a pay wall, they do not index it, and that’s always been their policy. So, Expert Exchange has always been trying to kind of flirt with that.

Andrew: So, the other thing you said earlier was you, guys, went from zero to $6.5 million in two years. I watched that growth on Hacker News, I watched it with the charts on It just seemed like questions and answers were just an easy way to grow a community. When you, guys, open up the platform to other people, I created my own little question and answer site. It grows faster than a blog, definitely, faster than a blog, but zero to $6.5 million in two years, I’m not on the way to do that, doing that with my community. Why did you, guys, grow so fast?

Interviewee: Having good quality answers, I guess, just getting good answers. There’s a few reasons why we have good quality answers and those go back to the design of the software itself. So, it’s things like the fact that you’ve been vote on which answers are the best. The fact that there’s a reputation system that prevents people from spamming and let’s you know what the good answers are. The fact that it’s a wiki so that every question and answer can be edited so to improve and [xx] other features there. It’s basically nine of them, if you look for a video of me getting a speech at Google about Stack Overflow.

If you search for that on the Internet, they’ll find my explanation of the nine key principles. Part of that was just brining in that critical mass to the audience. So that’s a theme I will return to later in this interview. We brought in this huge mass of people who rejoin with software, and Jeff Atwood brought in the people who recoding are on the first day to make sure it’d be a pretty large mass. You probably brought a lot of Mixergy people into your site, and that created sort of the initial audience. Eventually, they’ll start asking questions, getting answers, and Google finds our stuff right away and then somebody types the same question to Google. They find the answer and they discover your site and continues to kind of grow like a snowball in that way.

Andrew: Let’s talk then about the news. Can you say it now?

Interviewee: Sure. So, obviously, we started out with programming and our only ambition was to make it easy for programmers to get answers to their questions. We slightly expanded it at the seams a little bit by adding Server Fault, which is a site for systems administrators. It’s very closely related, and Super User, which is the place for us to put all those computer questions that didn’t really fit on our programming site or on our system administration site. So, those are the sites that we’ve been calling the trilogy, three sites that we made that basically, are meant to get good quality information out there for programmers.

At some point, we realized that we had accomplished our mission, and we look for the next thing to do. At the same time, both Jeff and I were frustrated that you would do a search on a programming question on the Internet, you’d find the answer on Stack Overflow, and it would be awesome. And then you’d do a search on your question about Siberian huskies or using your iPod to best effect or a cooking problem that you’re having or dentistry or whatever it was, and you find yourself in one of these BBS sites. It’s basically old styling news net quality, forums, discussion groups, where you, basically, when you get your search results, all of these is a conversation that people had. You start there, read the conversation, decide what the answer is, and try every single one of those purported solutions to your doggy’s diarrhea before you can actually can get to the true answer.

Stack Overflow would explore that and Stack Overflow, if you ask, “Why is my doggy had diarrhea?” the number one answer would be the reason will tell you exactly what to do. You’re on at the beginning of a research project or you’re at the end. So, we realized that this platform had applicability to so many other domains where people are looking for objective, factual answers to their specific questions. They want expert answers, they don’t want other people, they don’t want the teenage girls that use Yahoo answers as a chat environment…

They don’t want other people, they don’t want the teenage girls that use yahoo answers as a chat environment not really for answers or questions, but just to chat. They are looking for experts talking amongst themselves. So we decided to expand the mission a little bit. Our first effort to do that was called StackExchange and now I am referring to it as (((TOO FAST))) and the idea was hey, we got this great software that makes communities really work, let’s sell it. We decided to sell it as a hosted service starting at $129 a month where folks like you would come along and say, “Hey, I’ve got an idea, I’ve got an audience, and I have $129 a month. Why don’t you sell me the software and I’ll build this awesome site.” And that’s where we were and that’s been in beta now since the beginning of September. A nice guy we hired here at Fallcreek software, Aaron Menpaw started the development on that. He took the stack overflow source code, and modified it so it could be white labeled so you can customize the look and feel, and create your own login mechanism and make it look like the rest of your website. We have been in beta for 6 months. Over the course of the 6 months we have found a lot of successful sites have gotten created, my favorite is for PhD level mathematicians. If you go to this is not algebra questions from teenagers, this is PhD level or post doc level mathematics. It’s a really awesome site, get’s a lot of traffic, (((TOO FAST))) It’s an incredible resource for mathematicians that basically erupted on the StackExchange platform in about 2 months, that it went from 0 to pretty much all the mathematicians on the internet are on there right now. So there were some really exemplary examples of great Q&A sites that popped up on the StackExchange model, but there were so many other sites that opened and just became ghost towns. For every mathoverflow there are 20 sites that have less than 3 users. Basically the founder and their cat are on there asking each other questions and it’s just not working. What we realized essentially was that our StackExchange model required you to have three things to make a community: you had to have an idea, you had to have an audience, and you had to have $129 a month. We weren’t creating enough communities, the communities weren’t high enough quality and they weren’t getting audiences. We kind of went back to the drawing board before StackExchange came out of beta, even before we charged anybody a penny, we went back to the drawing board. The stack overflow team came out here to New York where the StackExchange team resides. We spent a whole week discussing what our strategy was, what our goals were, what our corporate visions were. We came up with a specific corporate goal saying, “Making the internet a better place to get expert answers to your questions” so there are a few keywords in there: one is, our goal is to make the internet a better place to get answers. It’s not to own those answers or to own anything, we are just trying to make the internet a better place to get answers to expert questions. This isn’t like, “How do I be more popular?” there is no expertise involved in how to be more popular and that’s the questions those guys can go to yahoo answers. It’s not, “how do I get lipstick out of my shirt?” You don’t need that much expertise for that, there’s a million places to find that on the internet. It’s the deep, dark, difficult programming questions, tax questions, and Siberian husky puppy questions that you really need an expert to answer. That is the new goal of the combined stack overflow, stack exchange team. Here’s the change, and this is the announcement we are making today, number one: we don’t want to limit ourselves to sites from people who can pay $129 a month. Because what we were doing is getting the entrepreneurial people that wanted to create sites and maybe try to make a business out of it. We weren’t getting the people that had a great idea for a site that maybe didn’t want to pay for it. So the first thing we did was raised some venture capital, and I’m not going to announce that today that announcement is forthcoming, but we raised some venture capital that gave us the confidence to say you know what stack exchange is free. We are not going to charge for it, we are never going to charge for it. The second change we are making we looked at the kind of stack exchange that were getting setup and we didn’t want a million ghost towns. The biggest problem with stack exchange right now is that 3 people might setup a site to talk about sequel server, 3 people might setup a site to talk about start ups, and 3 people might setup a site to talk about Siberian husky puppies. While I’m all in favor of let a thousand flowers bloom, when you do that with these sites that need critical mass that need network effects. The results is you get 3 small sites, none of which ever get to critical mass or never get the right network effects. These sites could be much much more powerful if there was one question and answer site for start ups, sequel server questions and for Siberian husky puppies or maybe

minute 15 to minute 20

Interviewee: …Siberian husky puppies or maybe Siberian huskies, in general. So, what we decided is that rather than just letting anybody create a site by just typing in a title and creating a site, that we would create a democratic community-driven process to decide which sites to create. Rather than letting anybody create a site that would then wind up in a ghost town, we try to best serve the community to find out if they could provide us with enough of an audience to make a site really get off the ground.

So, basically, the second thing that we’re announcing, besides it’s actually just going to be free, is that the new Stack Exchange sites will get created through a community-driven process of discussion, a bunch of people talking about what other sites they might go to. We’re going to start with the existing audience of Stack Exchange, Stack Overflow, Server Fault, Super User, those people we got the audience. Rather than trying to create new communities from scratch, we’re going to go with that audience and say, “What else do you want to talk about? What else are you experts in? What else should we open a site about?” And let them decide.

Let’s say somebody proposes Siberian huskies, the next thing we’re going to do is have a commitment process, probably a 1-week long process. People come and sign up and say, “Yes, I would go to that site if it were to exist.” We’ll try to see what level of commitment we’d get. That commitment, we’re going to actually look at the people, the actual people that say that they’re going to commit to that site. What their reputation is on Stack Overflow, Server Fault, Super User, and all of the other existing sites. What kind of contribution they made to those communities to try to judge if we’ve got enough critical mass to open a new community.

If we hit that threshold of what we think is enough critical mass to create a new community, we’ll create that [xx] on for two or three months. If at the end of three months, it still has a lot of traffic, it’ll become a permanent community that get its own top level domain name. We’ll get a graphic designer to design an awesome template for it and the site becomes permanent. If it never gets [xx] from us, we’re going to close it, because apart of making the Internet a better place to get questions, it’s not to leave around these ghost towns, where most likely you’re going to a question but you’re never going to get an answer. We’re going to tidy up and we’re only going to leave around the site where people can really get answers. So that’s a whole lot and I just sort of crammed it to one big long speech. Let me pause, take a deep breath, to give you chance to ask any questions you might have.

Andrew: Let me through that I understand this first of all before I ask questions about it. The way you store(?) is, anyone who like your question and answer site can go and create their own, kind of like with Postares(?). You can create your own Postares and you can either leave it at your or you can have it at your own domain. That was the way it was before, except you, guys, charge $125, I think, it was. What you’re saying now is all the people who did that, “That’s not the way we’re going to continue to work. You, guys, need to unwind your sites.” True?

Interviewee: Wait. We’ve created a new process for our future sites. So, I should clarify that the legacy of Stack Exchange sites like yours that were already created, we’re going to, basically, work with you to either transition or to continue or remove off the platform. We have a whole complicated program. So, basically, we’re trying to make the Internet a better place to get answers to questions, [xx] launched Stack Exchange 1.0. Like a great example would be and a bunch of others where people are getting answers. We are not going to touch those sites, they continue on to the existing terms except that we won’t charge for them. We’re not going to break any sites that already exist because that will be making the Internet a worse place to get expert answers to your questions.

On the other hand, going forward, if you want to make a new site about sewing or couch potatoes or regular potatoes or Idaho potatoes, in going forward, you will have to go through this community process and prove that there’s a critical mass of people that want to talk about Idaho couch potatoes. In which case, we will create the site for free, otherwise, we won’t. So, I want to distinguish between sort of legacy sites, which are going to be grandfather under a fairly complicated system, which will surely be up on our blog. Let’s see if it’s showing up there yet. Yes. Go look at You’ll see that we, basically, got a system to the grandfather sites and we’ve got, basically, two community managers already standing by to talk to you about the transition if you’ve already built a Stack Exchange site. Those are Cornell(?) and Robert.

Andrew: I get that. I called up when I first got the email from you, guys, saying, “Hey, we’re about to blog this.” That’s very generous of you to give us the heads up and to give us an opportunity to call in and ask any questions, which I did. I was told that you, guys, won’t really be developing my version of Stack Exchange, which is the legacy version now, if that’s what it’s going to be called, but you’re also not going to be discontinuing it. Just continue if you want, or if you want, we can also take our content and go to a different platform.

Interviewee: Right. Let me just talk about the history of the source code.

Interviewee: Let me just talk about the history of the source code, which is about eight months ago. We forged Stack Overflow and Stack Exchange. Probably a mistake in retrospect, but not worth going into. And so Stack Exchange has a bunch of enhancements, and Stack Overflow has a bunch of different enhancements. And that’s the current status. Now for Legacy Sites, we’re probably going to give you a choice of staying on the old platform, or one more. So you can kind of get some of the best of both worlds. It’s sort of, the truth is, to be completely honest, it depends on the site. There’s a lot of stack exchanges out there. Not yours, but I’ve seen stack exchanges out there that just have no traffic. They get, you know, one or two visitors a day. Honestly, we’ll keep them up for three months. We’ll give them access to their data. We’ll give them help and advice migrating to other platforms that give them a similar functionality. But we’re not going to work that hard on supporting sites that only have three people going there a week. On the other hand, for the sites where people are actually getting answers, we do sort of want to cradle them. In some of those cases, people being perfectly happy to take their site and essentially give it back to the community, make it stop being kind of an independent business, and move it into the network of stack overflow sites. And there will probably be some examples of that. Next thing. That’s one thing we can do, and then we can move you onto the new platform. The other thing we can do, as long as the site is getting good people use, is getting people good answers to their difficult questions, you know we’ll try to keep it open as long as we possibly can. And that means pretty much, I don’t want to say indefinitely, and I don’t want to say that we’ll never compete against it, or somebody else won’t make another site to compete against it. But, like I say, we try to keep our promises that we made in the past. And we don’t want to damage communities.

Andrew: How many active stack exchange sites are there?

Interviewee: There are, depending on how you define active, I would say there are probably between 20 and 40, where people do actually get answers to their questions, if you use that as a metric. So one metric we were using is, do you have ten users that come to your site every day. That’s about the minimum number of human beings that have to visit a site, and actually be able to get people answers to their questions. There are probably about a dozen that are incredibly active, where you get instantaneous answers. They’re really awesome. You know, there’s another couple dozen where you’re likely to get a decent answer eventually. The rest are all just ghost towns.

Andrew: Wow! So only 20 to 40 sites with ten active users. I really expected it to be bigger. And it’s not bigger because you’re saying there aren’t enough people out there with a big enough community to seed these question and answer sites.

Interviewee: Right. Well, they need the coincidence of a big community to seed it, and they’re willing to pay money for this thing in the past, because we were charging money for it. And so they’ve got the audience, the idea, the money. And those were the three things you needed. And we’re kind of eliminating all of those problems by saying it’s going to be free, the ideas are going to come from the community, and we’re only going to create sites where we already have people talking. Right? So rather than saying, “I’ll bet there’s a lot of people that would like to talk about auto mechanics”, we’re going to go to our community. And today it’s a lot of programmers. And we’re going to say, “You know, what else are you guys interested in? I’ll bet you a lot of you are interested in talking about personal finance, and auto mechanics, and dating, and graphics cards, and gadgets. And there’s a whole bunch of topics that it’s obvious there’s going to be a lot of programmers that want to talk about. And over time, what I expect will happen is that the programmers will spawn new communities. Those communities will eventually develop their own audience of people who aren’t programmers. Let’s say the personal finance site. Right? There’s a lot of programmers that care about personal finance. They’ll create the personal finance site. It’ll be awesome. It’ll start to attract all kinds of people from all over the place. Suddenly the personal finance site will be its own big, growing thing, with a big audience of people that maybe some subset of them are interested in fashion. And they’re going to create the fashion site. So eventually, we think that we can spread to a lot of other topics, by starting with our audience. But the most important thing is not to just go out in the middle of nowhere, and just say, “Hey, come look at me. Make us time.” But actually to take the core audience that we have, and to continue to grow it, rather than just try to create new audiences.

Andrew: So why don’t you go fremium? If money’s an issue for your users, why not just allow anyone to create a site? And as long as they have under a thousand users a day, they can use it for free. And once they get big enough, then they can start paying.

Interviewee: We sort of had a staged system. I mean it wasn’t quite fremium. But when you do something like that, people look at that and they say, “Hmm, you’re going to penalize me for being successful. I’ll make this site, but I’m afraid that it works. And if it works, suddenly somehow I have to come up with this $129. And I can just slap some Google Adsense on there, but I don’t know if that’s going to work. And it’s not, you know. This is not. I don’t think that… I think that that model, to us, makes a lot of sense because we’re both entrepreneurs.

interviewee: to us makes a lot of sense because we’re both entrepreneurs and we think, hey lots of people think like entrepreneurs. There are also people out there that are just programmers in a company somewhere, who are real experts on the art of raising Siberian husky puppies and they have raised 64, and they participate in ididarad and all that kind of stuff, and they want to talk about that and they don’t want to think about the money and that’s just going to immediately say oh never mind I might have to pay, or it might become big.

Andrew: Why not just keep it free and let those guys create all the sites they want and have your own advertising in there?

interviewee: We have actually been proposing that and we have talked a little bit to our communities about hey, we will bring you advertising. The weakness there is that unless you have some kind of process to ensure that each community only gets created once, you are likely to get a lot of communities on the same subject

Andrew: But aren’t there vBulletin boards on the same subject? I know that when I do a search for a solution to a Mac question I have, I come across nothing but vBulletin sites and I can’t understand how they all survive until I dig in deep and I see each one has its own character, its own approach to answers, it’s own community that’s talking about getting together on the weekend, or let’s see pictures of your dog. Why doesn’t that work that way for Q&A sites?

interviewee: I think that’s awesome, I love those communities, I want those communities to continue to exist, those communities are absolutely terrific. Let me narrow it down, not Q&A in general, let’s look at expert Q&A. Let me just go to StackOverflow. and I’m looking at a bunch of questions on the front page that have been recently answered, and most of them which have good high quality answers have been viewed less than a 100 times. Some of them have been viewed like 5, 10, 15 times. The reason is that we’re trying to get, we have probably 600,000 questions right now on stack overflow, the reason we have so many questions is that these questions are very narrow, very specific. That’s why they are getting 50 views, 100 views each over their lifetime of that question. 100 people that are going to look at that, But it’s the right 100 people. It’s the 100 people that are having that exact very specific problem. Like I live on west 74th street and my puppy keeps eating something in the street that is giving him diarrhea. I’m going to find the other 3 people who walk their dogs on west 74th street and one of them might now that there is rat poison there that the city puts down. To find those 3 people, that is sort of an example of the extreme long tail question and answers which are expert questions and that is what we are trying to solve. In order to solve that you need (((TOO FAST))) or whatever that’s fine for asking the same questions that everyone is asking, or the easy questions like, what is an easy startup question? Should I be an LLC or C Corp? You will find 800 answers to that on the internet. The real question is, I was organized as an LLC until April, and then April I switched to a C Corp, and it’s a (((TOO FAST))) corporation and I want to know when I have to send the K1 to the people who are members of my LLC…You know questions like that. You start to get really narrow. Somewhere out there 12 people have the same problem and one of them did a heck of a lot of research to find the answer and can tell you what to do, and you have to find that 1 person out of 12 and for that you need really big sites and that’s the reason why having 100 gardens bloom is not going to get you answers to those questions.

Andrew: I see a couple people in the audience who have some questions about what’s happening, why don’t we explain it now. See if I understand this right now. There are 20 to 40 sites that are actively using the question and answer platform that is stack exchange, those 20 to 40 sites lets hold on the side. From now on if you want to create a community on this question and answer platform you can’t just go off and do it on your own, having money isn’t enough to get you to do it, having your own community isn’t enough, being able to buy your own domain isn’t enough to do it. If you want to create your own community you have to go within the stack overflow community and say, “Hey guys why don’t we all talk about this one subject. Let’s all talk about the iPad. Who has questions about the iPad? Who wants to talk about and answer questions about the iPad? And if there is enough interest in doing this then you guys will create that site about the iPad, you will run it out there for a few months to test it. If the demand holds up you will continue if it doesn’t you’ll say hey, you guys aren’t really that interested in it we’ll pull back and try something else. That’s the way it’s going to work in the future. These 20-40 sites what they are telling us is, contact them.

Interviewee: There gonna work with you.I I called them up myself. I said,” whats going to happen to my site? Is it going to go down?” they said,” No you can keep running it as much as you want. We’ll help with critical updates but we’re not going to make this into the next Mahalo answers or yahoo answers for you. We’re just going to help you, support you, were going to be there. You trusted us and you trusted a good company.”And really I gotta tell you

Joel: Hey Hey you should come work for us

Interviewee: I I actually was really upset when I got that email, and confused more then anything else. I read the PDF and I never read anything that comes in my email thats more then four sentences. I read the PDF I still was a little confused. So let me try calling them up. The guy didn’t pick up the phone. I sent an email saying ” Hey I tried calling, you weren’t there.” I get a nice email, I get a nice call right back. The guy walks me through and made sure I understand it all. I’m used to thinking of you as the company Joel. It is so nice to see that other people in the company are live up to the expectation that you set for us.

Joel: ya well we got two we got two

actually two people that are literaly well that is there primary responsibility right now. Robert Cartano who many of you know from the stack overflow community from mega sites is has been highered as a community manager full time working for stack overflow / stack exchange and oh here he’s actually being sort of a communitee manager of the VA’s

Interviewee: The video just suddenly sped up as you were saying that. Let me say this to you. I’m very understanding and I really appreciate how great you guys are. I don’t know that this is the the right decision or not to make. I’m not here to say. “Hey Joel made a great decision here.” Its to early I think to say. You guys are just gonna keep feeling your way through this business until you find out what the perfect direction is. I wonder though. How much of this is you and your company not really getting that consumer market. You’re dead on with your audience of teckys but when it comes to even a site thats built for The general audience of entrepreneurs. They don’t understand what open idea is. They understand what facebook login is. they don’t understand about the difference between a comment underneath a question and answering the question on the bottom. It feels like its just not in your DNA. How much of that is true?

Joel: Let me Let me take those one at a time.The Number one thing about Open ID we’ve recognized that has always been the number one fixture request at stack exchange. The first things that we implemented was an alternate log on mechanism for the stack exchange sites. Other then open ID so that’s

definitely something that we’ve known from day one would be necessary the day we moved out of the programmer ghetto, and the second thing that you mentioned was wait, wait

Interviewee: Oh that it’s not in your DNA to talk to an audience thats not developers, thats not geeks

Joel: No but you had a specific thing that you said. The open ID

Interviewee: that I talked about?What do you guys say in the audience?

Joel: Gosh darnit, somebody remind me of what you said?

Intreviewee: Alright somebody in the audience if I thought we hit it. If anyone in the audience caught it help me understand. I was asking about my question. Maybe this will jog your memory is

Joel: Comments verses answers. Thats what it was.which is people come to the site and that is a big problem on stack exchange right now. You look at these other sites. Theres a great one called tax queries. Doesn’t

quite have critical mass I would say but it does have a few reasonable people that know somthing about tax law, and you’ll see that people don’t really know how to use the comments verses the answers so much on that site, and we’re specifically addressing that with a. Rather then going out to the tax lawyers and making a site for them and hoping that they get it. We’re starting with a community of people that already know it.So when we. When somebody proposes a new stack exchange site were not going to open it until we know that there’s a critical mass audience of people who have already used stack overflow, server fault, and super user, and we’re going to look at the kind of badges they have on server faults and overflows and were going to look at the reputation they have on those three sites, and try to figure out. Hey is there a critical mass of this new site of people that get how the software works.They know how to tag. They know what tags should be used for. They know the difference between a comment and an answer. They know how to use the answers. They know to vote for things. They know to vote for questions not just answers, and all that kind of stuff. And thats the kinda stuff that are a problem on stack exchange and I think that problem will go away when we start building new sites with a community that already gets the software certainly during the beta faze. So one part of this new process is the site created.They get seated during closed data by commenters who commented to the early site before there open to the general public. So in all these areas your absolutely right. It’s not in our DNA and if you look at the popularity of sites. You’ve, go to super Stack overflow right here and you’ve got server fault let me aim my camera a little lower here, and you’ve got super user also growing pretty fast. But the further you get away from that tentpoll of programmers which is my audience and Jeff Atwoods audience the less draw we have and the less ability we have to draw


Interviewee:unless you understand it.And so [xx] going out and hoping to create a site for [xx] and Husky fans.We are going to try and create new sites where the existing audience will talk about the other stuff.And hope that in that way it will look into the rest of the world.

Andrew: People were asking here in the comments, what they were saying, Andrew [xx] were also built on stack exchanges.How many programmers in tax accountants are the same person?I think what they are all saying here is, if they are not enough tax accountants within his current community,he is not just going to create the tax accountants site.And source is saying, is it limiting the way to start a site.Source is saying essentially with mouth overflow site had been launched this week.Even though it is the biggest site,i think what you are saying is it may not be,even though it is the biggest site,they may not have to create that because it is not just in the DNA to create it.Possible?

Interviewee:Let me take both those examples.[xx]not that i don’t know anything about it.Everytime i went to BCs and i pitched the company,i showed them [xx] and proved that this wasn’t the software that only worked for programmers.And i said, look this is [xx] to me but there are no objective [xx] and style and pointed to the question, it was like ,what goes well with the following [xx].And i said [laughs] those are not objective questions,those are not objective at all.That’s my style and that’s should work.But in style [xx]there is often an example.I don’t think the site would have been created as quicker.I think that might have been a third generation site.So the programmers might create the gadget site and the gadget lovers might create the consumer site,consumers might create the shopping site and the shoppers might create the fashion site and so on and so forth.And so you can’t [xx]went to the other, that might take several generations.In the case of math over flow,that was a big [xx] inspiration.I talked to Antoine who created that site and long before we had stack exchange,he was constantly trying to contact us and beg us to make a site about math.He though that was the greatest idea.And what he did do was he went out and gathered a critical mass of mathematicians that he knew will participate in the site and make sure to make them all jump into the site on the first day.So they talked to each other.And by having that critical group there at the same time on the first day when he emailed them all.There was only thirty people in the mailing list.By having the thirty come at the same time, he immediately answered each others questions and created a snow ball effect of critical mass.And i looked around and said,”wow,this is working”.And he immediately,some of them were bloggers and started blogging about it.And the site sort of took off and really resumed.And so,under a new model the same thing would happen.He would go on to that group and say,”Hey,How many of you are mathematicians?”.Probably a lot of programmers are mathematicians.Do we have enough people here?If [xx] we will look.We will see that there are thirty committed people.We will see that the[xx] have the skills to make it work and we probably would have welcomed it.

Andrew:I think, i seen how it works.If you guys have any questions about the process in the audience,please ask Moses.I know you have questions.I missed it in the chat, please ask it again. I just covered you because the person who i looked to for guidance in business,[xx], said he looks to you when it comes to development.Joan on software,Joan Polski has been the teacher and the leader here.How does it feel to have to admit publicly that the direction that you went in, wasn’t the right direction?

Interviewee:The number of mistakes that i made just in Fall Creek.We created products that did not work,products that were sort of underwhelming, we created products that absolutely went no where.My absolute favorite product that i created was the Indian version of the Joan software program job listing board.We made fifty rupees after that with two job listings for programming jobs in India.That was probably the biggest and the [xx] failure.[laughs]I was very excited about that.Its going to be huge [xx][laughs] trying to get jobs.We only charged twenty five rupees instead of what we are charged now,that is, 350 US. SO who bought that, the number of mistakes [xx].I got business, part of it is to find the right model and you need to travel and figure out and see whats failing and you kind of tackle in different directions. I think in this case, we are kind of lucky, we were able to figure out that it was not working and changed the model before we had taken a penny from people.And so the truth is,Some people are going to feel like that we have moved their cheese but at least they are paying customers.

Andrew:What do you need funding for this business model,if its going to be slow growth,if its going to be organic.If its going to be using the technologies that you have.And speaking to the audience that you already have why do you need funding?

Interviewee:Well, the first thing is we don’t want to limit our growth by the amount of cash that we have in the bank.

interviewee; ….and we also have a belief that the share with our investors that and you can look at that some other similar companies like twitter, face book,and google… where the ideas was lets make it big and i know it sounds like dot com one out thing we know that we can make it big and organic revenue model that will emerge that will pay for this thing, that it is not expensive but you have to make it big first.This is the kind of things that we can only do when in the and all the mathematicians, all the programers all have competitions.

just to give you an example that you have a fashion website just about guys about you go to that advertiser for advertising they are not going to talk to you they are never heard of you they are not going to advertise it the best you going to get is some add sentence and that one sentence is a big waste of time and its going to be all wrong that are really texted ads really related to time and page and content and its very hard to marketise that and on the other hand if you have a website some of the property top five style slash destination on the internet then your ad sense your advertising model much more robust and you have for more opportunity what happens and i am not saying advertising but happens in the advertising world that the national advertisers really at the place it heard of and so you have

Andres : organically, you can develop science of advance with more formalities and lets you thing buying and final ways to bringing

mturk april 13 2010

transcribe min 45 to 50

Interviewee: …in all kinds of different ways. Some of them are awesome. I’m a big fan of Jason Kalkannis. Um, I can’t really speak to what they’re doing — what their particular strategy is. I do think that we have the fastest growing Q&A site, uh, and the only one that gets detailed, deep, technical answers on the very very difficult, long long-tail questions that only 40 people in the world care about. We’re the only site there that answers those questions.

Andrew: Uh, let’s see. Okay, so Moses in the audience, he has a stack exchange site. It’s called and he’s asked me a couple of times to ask you, What’s going to happen to his site? “How will it affect sites like mine? My goal is to build a community, hopefully will convert into paying customers.”

Interviewee: Right. Uh, well, before you — I don’t — I’m not familiar with that site. Uh, there is a blog post up right now — I If you scroll down to the FAQ, there’s very detailed specifics about what’s going to happen to various sites and like I said earlier, if that doesn’t immediately answer your question or if that answers it in a way that doesn’t make you happy, we’ve got two people to talk to and we’ll discuss, you know, what to do going forward.

Andrew: What about We know that they’ve invested, seems they’ve invested hundreds of thousands of dollars in this platform. They’re really building it up and building a business on top of it, and they want to know that they’re — I talked to the founder. He wants, he wanted at the time when he started the site, to know that in the future there’d be Facebook integration. He had all kinds of ideas for the site. [Yeah] Now none of those are going to happen.

Interviewee: Uh, well, they might if he moves his site onto the new platform.

Andrew: But then he can’t control his own advertising, right?

Interviewee: Uh, well, like I say, he should contact us. [Okay] That’s basically what we’re — yeah, look, we’re working, we’re working with people. What I’ll repeat and I say this a few times in the FAQ, at sites where people are getting good answers to their questions, um, we don’t want to do anything to break those sites. Those are … is not to own it, …

Andrew: But the fact that you got …

Interviewee: … as much of the Q&A on the Internet as possible —

Andrew: But the fact that you’ve got — sorry, there’s a lag, I’m sorry. I hate Skype for these issues. I wish I was in your office, Joel, and we just … [Right] would be able to sit and talk like human beings, [laughs] not like — I don’t know what this is, CB radio truck drivers. Um, but the fact that you guys aren’t going to be building this out does put his business in an uncomfortable position. [Yeah] True?

Joel: Well, I mean, you could — whenever you (and I’ve been saying this to people since the beginning of time) [indistinguishable] on a piece of technology that you bought or from somewhere else, uh, you really should make sure that that does what you need. And if it does what you need, go ahead and use it. If it’s not, doesn’t do what you need today, uh, then, you know, you are imagining that one day it might have Facebook integration or something like that and you’re going to build a business, on the hope that some other businessmen some day decides that it’s a good idea to implement that Facebook integration — well, there’s a weakness in your business model right there. I don’t want to say that there is in his, but the point that I really want to come back to is that our corporate goal is very very specific. It’s to make the Internet a better place to get questions answered. Expert answers to your questions. And that means that if there is some community out there, like Siberian Huskies, that is using The Bulletin, or PHBBB or one of those awful old-school technologies, to get people answers to their questions, it’s just not working. And we want to replace that with something better. However, if there is a community — a good example might by, which is using not our software but a clone of our software, and is actually getting people good answers to their questions — we don’t have the goal of stealing their business or taking over or owning that particular segment of content. Our goal is to get answers [indist] startup space, which is your stack exchange, Darnish and Jason on There’s already three of those sites. I wish there was one, but there’s three of them and frankly, that’s the last area that we need to go into to make a fourth stack exchange site. And this is not a corporate mission. Our corporate mission is to make sure people get answers. We’re going to put our effort into other communities that have asked for it. I think that what would happen if the community said “Hey, what do you want to talk about?” and a bunch of people said, “Startups!” there would be a lot of people on there saying, “Hey, there’s already a great startups stack exchange. There’s no need to make another one.

Andrew: Guys, if you have any questions put them in the comments. Please, make them comments not statements, because it’s hard for me to read your statements and rephrase them as questions. I don’t want to just give him your points or understand your points here now, live.

Joel: I will read them immediately after this video.

Andrew: Moses is asking, Who survives?, or I never thought of my site as being in that startup space. I wanted a place for us to ask questions. I said, you know what? I really like the conversation that’s going on on Joel’s sites. …

Andrew: Let’s see if we can do it in a question and answer – if we can have our community in a question and answer format, instead of a bunch of people making statements at each other. Let’s see here – Joel, it does raise an issue here, with the cloud: I used ethernet for my transcripts, ethernet was sold to Google…

Interviewee: Etherpad.

Andrew: Etherpad, right, right, ethernet is the cable I’m talking to you on. It was sold to Google, I can’t create new pads on it, after two days from now, something like that, it’s gone. What happened recently? Someone creates an app on top of Facebook, Facebook decides to adjust what goes into the status feeds, the business is gone. We’re all being transitioned into the cloud, but I feel like some of us are going to get tossed off the cloud and onto our face! This is a tough place to build our environment on people’s platforms!

Interviewee: Ah, oh boy, you nailed it.

Andrew: How do we protect ourselves? Is there a way to protect ourselves?

Interviewee: It sort of depends. And this is something, actually, there was a discussion here, and I’m kind of getting off the reservation here, not really thinking about this from the stack overflow point of view, but Fred Wilson wrote a blog post recently about Twitter, and then it turned out that they were announcing something that he must’ve had in his mind… but basically, if you were a platform developer, and it is obvious that- sorry, if you were a small fry, and you were looking at these platformm developers, and you say, “They have this hole in their plaform, I’m gonna go fill that hole.”, you’re in a dangerous position. You are snatching nickels out of… out of in front of an oncoming bulldozer. You’re going to get a few nickels, but sooner or later, you’re going to get bulldozed. Because if Twitter is going to be successful, it has to have an iPhone client. It’s just inevitable. It has to have URL shortening. It has to have the ability to attach pictures. That stuff is just such an inevitable part of the Twitter roadmap; the fact that they couldn’t build it right away – well, it’s great that they have an API, but they’re gonna do it. Go build something different! Maybe do something with their platform that adds value to it, rather than something that tries to steal a little value from them while adding value to them? ’cause otherwise, what you’re looking at is a very very risky business proposition.

Andrew: Yeah. I like the bulldozer analogy! It’s what it feels like! You’re grabbing nickels, from an oncoming bulldozer, Facebook will eventually create its own virtual currency, it’ll create its own OfferPal, and all these other business are just going to have to deal with it, it seems. And a situation like yours, it’s not the same – people are asking in the audience, is that what you’re saying about Stack Exchange? – no, it’s not the same at all, it just didn’t work out. I think it’s more similar to Etherpad. Etherpad tried an approach, the approach didn’t work, those of us who were on it needed to understand that it was new technology and there were risks. And maybe that’s the answer-

Interviewee: Yeah.

Andrew: -to just understand; it’s great to build startups, but sometimes you just can’t build your businesses on top of startups. They’re just tring to figure things out – think of all the people who were using Seesmic in their comments and suddenly you just don’t have it! Seesmic video in their comments.

Interviewee: Right, right.

Andrew: Okay. I think I’ve gotten to all the questions from the audience. Joel, any last word from you? Do you miss blogging, Joel, do you want to come back to MixerG and blog once a week with me? You do video like this, you’re perfect at it! How about five times a week?

Interviewee: No, no. *laughs* I don’t… you know what, someone came up to me this morning, if you’re watching this, this is even worse, but somebody came up to me this morning as I was walking my dog, and said, “Excuse me, are you Joel?” and I said, “Yeah” – this happens to me all the time – and he says, “I recognized your dog. That’s Taco, right?” So now my dog is more famous than I am. Anyhow, no, I do not miss having random strangers walk up to me and recognize my dog, and say “That must be Joel Spetzky, famous tech blogger, because I recognize his puppy!”

Andrew: We can’t get a headset on your dog so he’s not welcome on MixerG, you and I would do once a week conversations with Joel, it’ll be “Joel on video” – somebody get that domain name. We’ll figure it out..

Interviewee: The only thing I want to mention is, if you do have specific questions,, we’ll be all over it, announcing them and if you made a stackexchange site, like I said, just contact us! It’s not so bad, it’s not terrible. Bad things won’t happen; we’re not going to turn anything off tomorrow. There’s stuff that we can leave running indefinitely. If you really want – we’re not making any promises that we can, we never made any promises, you know, I can’t tell you right now, “I’m going to continue to develop a product that has no market for the next, you know, unforeseen future. Adding new features to it that only sixteen people really want.”

Andrew: What do you recommend that people go over to? Should they go over to Sponge at, or, that people are saying in the audience, I know people at HackerNews built a clone of your site, is there one that you recommend people use? If they want to transition?

Interviewee: I don’t know anything about that, probably by the time we start asking people to transition, we will have figured out which one of us has emerged as a good alternative that we can honestly say we recommend; we don’t see ourselves as being in the software business, there are open-source implementations and not-open-source implementations and clones…

Interviewee:clones none of them are as good as what we have today they’re all kind of behind, but I think by the time we start asking people to transition one of those will have emerged and if we’re actually saying “Hey, get off and go somewhere else because we can’t run your server.” I think there will be a reasonable alternative at that point.

Andrew: Ok. And just to be clear, you’re not saying that you’re going to say that anytime soon. You’re not telling people who are listening to this that you’re going to kick them off your server soon. Let me continue. Presumably, if you’ve got one of these smaller sites you don’t have much action anyway and somebody needed to say, “Hey, it’s time to end the site. And you were thinking of doing it anyway and if they were going to charge you you were going to end the site anyway. This is maybe the kick that you needed.” Can I say that? Is that insensitive of me to say that?

Interviewee: Sure, and that’s what we expected. What we actually expected is of the hundreds of the site that have gotten created, hundreds of them are not going to care. They’re just going to say, “You know what, it wasn’t working. I’m going to close.” And some smaller percentage are going to say, “Hey, it was working and we’re going want to reach out to you to make sure that it continues to work.” And there’s some people in the middle and that’s why we’ll talk to you and negotiate and we’ll figure out what to do to keep you going.

Andrew: Joel, this would’ve been such a great interview if I what I would’ve done was spend… what now…57 minutes into the conversation if I would’ve said, “Joel! You screwed up! You screwed us!” And, you know, and fought you on this a little bit, but…

Interviewee: I thought you were a journalist!

Andrew: I’m too freaking curious about business and what goes on behind the scenes and I think that because of that I earn your trust and I got you even though there is so much going on today and I emailed you just before my last interview and said, “Do you mind coming on here, on Skype?” I know you feel and other entrepreneurs feel like this is a place to safely and fairly come on and talk about what’s going on and I’m glad and grateful to you for doing it last minute with so much going on. And I’m glad I could become a forum for that.

Interviewee: Sure! Thank you very much for having me on on short notice. It’s cool to eat up the scoop here on the new stack exchange plan. So, congratulations!

Andrew: I’m still getting traffic from that Juju interview. That juju may not have gone anywhere for the founders of the juju pad, but its doing great for me. People found me through that. Now who knows what this news is going to do for me?

Interviewee: Alright. Cool.

Andrew: Alright. Sorry, Joel. Where can people talk to you if they want to? Where can they find you? They all know your website. I don’t even know how to end this by giving you the last word.

Interviewee: Yeah, Joel on or, it’s about stack exchange.

Andrew: Alright, and I’ll see all you guys on stack overflow where I’m a member too. Bye.

This program was sponsored by

Wufoo– The easy way to add elegant forms and surveys to your site. (I use them on my site’s contact page. When we got married Olivia and I used Wufoo on our wedding web site to collect RSVPs because their forms are beautiful.)

Shopify – Thousands of stores are built using Shopify because it’s easy to set up and manage. Tim Ferriss recently announced a contest that offers $100,000 prize for the highest grossing store. Go start your store now.

Grasshopper – Entrepreneurs (like me) love and use Grasshopper because it offers all the features of the big, expensive phone systems (like multiple extensions, music on hold and call forwarding) but it works with any phone and starts at only $9.95 a month.

Who should we feature on Mixergy? Let us know who you think would make a great interviewee.