Podcast of Konstantin Guericke's talk on how LinkedIn Broke Through

Konstantin Guericke, LinkedIn Co-Founder

We loved hearing the story behind how the original LinkedIn idea was conceived and how they eventually managed to break through to mainstream. Co-founder Konstantin Guericke’s soapbox left all of us with some key lessons we wanted to share. Go ahead and listen to the podcast below as you read through the summary.

Listen to Konstantin's Podcast

iTunes RSSRead Transcript

Insight #1: The Early Days

Inspired by SixDegrees back in 1997, Konstantin bought the domain peoplemap.com to work on a vision of connecting people together online. At the time Geocities dominated much of this space, with everyone creating their own websites and trying to connect. Konstantin met Reid Hoffman, an entrepreneur with a pretty slick singles-matching algorithm running on a project called SocialNet, and together they recognized the potential for matching professionals in a similar way.

It was years later when Reid was ready to leave PayPal that the two of them sat down to give PeopleMap a shot. By this time they had noticed that as consultants, they got new business when an old client connected with a new lead, often over lunch, and referred them. The idea to give people a way to control these leads and stimulate strong word-of-mouth referrals online was where LinkedIn was born.

Insight #2: The Launch

Boy was it tough. Everyone hated LinkedIn. It was 2003 and Friendster was dominating social interaction on the web. Konstantin kept hearing from people: “Teenagers friending each other and trading pictures on social networks - sure! You want grownup professionals to friend each other over the web? This is too cute. Won’t work.” People also felt they lacked the first-mover advantage: “Even if it does work, Friendster will just add a business section to the site and you’ll be dead.”

Bloggers hated the idea of separating their private and professional lives in to and weren’t shy about letting everyone know their disdain for LinkedIn. PR was tough early on in this climate, but hard work and persistence in their vision paid off. Rafe Needlman of CNET was one of the first to give the company a glowing review and others began to follow.

Insight #3: Gaining Traction

When asked about the secret formula behind their success Konstantin said: “Not sure about our success formula, I know the fail formula for our competitors though”. At the time a number of competitors were rivaling LinkedIn among which were: Rise, Spoke, Plaxo, VisiblePath, ZeroDegrees and of course Friendster. A lot of the problems had to do with targeting the right market of users.

Most of the early traction was gained through PR, email marketing, testing the messaging on the site, and viral marketing. Konstantin mentioned that press can only give you a certain amount of traffic, to get to millions of users you need virality.

Insight #4: Design Mistakes

Mistake #1: The early LinkedIn version had a lot of privacy controls. So many that you could not see how many degrees away the person was from you. This was a crucial mistake since people did not see much value in the network if they coundn’t see if a particular person was 2 or 10 degrees away. The issue was quickly fixed in months following the launch.

Mistake #2: The early version of LinkedIn did not have an ability for people to search by a name. The LinkedIn team figured that people did not want to find someone they know but rather search by the title of a person they might want to meet. The issue quickly surfaced. The response was the introduction of Address Book uploader which allowed you to upload your address book and see who from your contacts is in the LinkedIn network. You could then send invites to folks you wanted to connect with. The search was also modified to allow name search capability.

Insight #5: Subscription vs. Advertising

Konstantin mentioned that if you have a website service which is free to use but has Gap t-shirt ads on it - it means that Gap has figured out how to get money from your website visitors but you have not. That’s kind of sad. Everyone’s goal should be to create something of value that people would pay for. Relying solely on advertising is a bad idea.


The discussion continued with Konstantin sharing more tips on getting consumers to pay a subscription fee, technological challenges to overcome, as well as what we can expect from the web in the future. We’d like to thank Konstantin for such a insightful and inspiring talk as we look forward to more!

Don't Miss Out on Our Next Soapbox

Soapbox transcript

Konstantin: The vision was to really combine three things. One was a homepage for professionals on the Web. I think it's a little ways off before we really delve into virtual reality for good. I don't think that's such a desirable thing anyway. It's typically what VCs do. They take something and say, "You've got to make it bigger."

Why would you want to search for a person by name? You know them. You don't need to search for them. They're probably not on here anyway.

Moderator : We're super excited to have Konstantin here. Of course, you know Konstantin, founder of LinkedIn, but even before that, just an amazing entrepreneurial story. He came over here from Germany, studied engineering, later on became VP of Marketing of Presenter.com which was acquired by WebEx. It was a presentation authoring tool.

Later on, VP of Marketing at Blackstone Interactive where he was marketing a 3D virtual reality chat system, and back in 1997 he met Reid Hoffman, started talking, had an idea for this social networking site that was completely different from everything else at the time.

Friendster was the big thing. MySpace was going to enter into space, people were focusing on teenagers creating videos at the time, and here it is a new idea for this completely new thing and we might even charge for it, versus everybody was doing advertising at the time. One thing led to another, lots of talking. LinkedIn launched at the time when afterwards MySpace entered in. It was just an amazing way how he gained traction and stayed stable and clean. I invited him here to just talk about those early days, how the idea was born and, of course, the traction that they gained. With that, let's welcome Konstantin Guericke.

Konstantin: Thanks, Moderator , for having me over. I don't get out here very often, but it's fun. I love the little downtown area. It's almost like you can't blink or you've gone right through it. I was trying to find parking across there and I said, "Oh, I can't park here. There's only 10 minute parking." Only later did I realize it's like a 10 hour parking. So very liberal parking policies here.

Audience Member: The funny thing about it is, the 10 hours is, they limit it from 8:00 to 6:00, 10 hours. So it's basically 24-hour parking.

Konstantin: Oh, I hadn't even thought about that. Somebody needs to take a picture of that. So, I thought I would share about, and we can keep it really informal, whatever you want to ask, anytime, interrupt and so forth. But share about the really early days, because that's sort of the least known and it's kind of fun, especially if either you have been entrepreneurs or you want to become entrepreneurs, I think it's interesting how it kind of comes together.

I think the first touch was right in, I think, in 1996-97, there was a site called Six Degrees. It became very popular. It had millions of users, but it got kind of a spammy reputation, but it sort of broke on the scene because it grew so quickly. People didn't know quite what to do with it. It was almost more like a reenactment of that experiment, the six degrees experiment, I forgot, some... You know, right? Well, now we'll know, the Stanley Milgram experiment. Thank you. Whenever I don't know, I'll ask Minal, because he was there pretty early too. What? Employee number what? First 15 or 20 or something like that?

I thought of something called, at the time, and I met with Reid and we talked about it, called People Map and I got actually the People Map domain. The vision was really to combine three things. One, was a homepage for professionals on the web, because at the time GeoCities was the big thing when it came to homepages. But I never made a GeoCities homepage and I'm in marketing. I knew the founder as well. I thought, "Why did I never create a homepage on GeoCities?" I just realized I couldn't afford to have a bad homepage. That is what a lot of marketing people, it keeps them away from doing things because it has to be at a certain level.

So that's why people just can do PowerPoint, because at least you don't screw up. You don't do anything great with PowerPoint, but at least you don't screw up then if you try to get a bitmap editor and try to do something with Paint or something like that. I remember for my wedding I used the Wedding Channel and created this gorgeous page with e- commerce and everything, but it was just templates. That's what I could do, is pick a template, fill out a questionnaire and then sort of magically it appears.

So that was going to be one of the three pillars. The second was self-updating address book, because it was always a pain to stay in touch with people and the third, an ability to meet new people through the existing people that you know. People Maps, the concept was that it would be like this. You be the center and you would see all your connections and little arrows and the arrows would mean what it is. Obviously, not like what it is today. And then each node would have other arrows and show the little bubbles around that and then if you click on that node and it would kind of move to the center. Xerox Park had a browser like that and sometimes in knowledge maps or something that technology is stilled used.

So we talked about that, but Reid and I met at the Avatars Conference, both involved in virtual worlds, kind of too early, I think it's still too early as Selenium Labs is figuring out now. So there is always this idea that always seems attractive but I think it's still a little ways off before we really delve into virtual reality for good. Maybe that's not such a desirable thing anyway.

So we were kind of leery because there was a lot of excitement. It was kind of like a little mini-bubble in 1996-97 just when the internet hit and we were already doing, like well 2D is passe and it's all 3D and lots of investment and oil crash and no revenue. So both Reid and I were like, "How are we going to make money with this little thing?" So he decided to go off and start a dating site called Relationships.com which later got renamed Social Net. Of course, the VCs saw it and say, "Oh, this is for dating and well that's too small of a market. If you'd use this matching algorithm you can also hook up people for work and for activity partners and for housemaids and all that stuff." It's typically what VCs do, right? They take something and say, "You've got to make it bigger." Of course, they all wish they'd invested in E-Harmony.

So we kind of went separate paths, but one of my friends, a guy named Jonathan Abrams, and you said I was LinkedIn, user number one, but it's actually not quite true. Jean-Luc, our VP of Engineering, is user number one, but I'm probably among one of the first ten. But I was almost more proud that I was user number 25 at Friendster. So if you look up my member ID, they started with, Jean's got 101 and I've got 125.

And Jonathan, we are friends. I knew him from his previous startup. He had a wonderful bookmarking and I was actually a power user. I loved bookmarks and organizing them, so I had like 6000 bookmarks and several hundred folders. Just perfect for his site. It was called Hot Links. He didn't know what to do so he sort of started Friendster without really planning on doing anything with it. But it sort of grew very quickly. He started reminding me, like, "Yeah, you know, what about this people map?"

In the meantime, what happened in the dot.com days, somebody just said, "Hey, this is a cool domain? Can I just buy it?" And I was like, "How much?" And he gave me a price and I was like, "Oh, sure." I haven't done anything with it so it was a great return of investment. I was a little sad later that I could have probably used that domain and actually make it come to fruition. But it kind of reminded me, and Reid and I sat down, he was ready to leave PayPal, so we kind of worked on that.

The other inspiration was really, I had been doing consulting at the time and someone asked me, "How do you get your clients?" I'm like, "Good question. I don't really know," because I didn't have a business card. I didn't have a website. So I listed all my clients and I was impressed how many I had. I didn't realize I had so many. Then I thought how did I meet each of them and I just always had a name of a person that introduced me. And I thought, "Well, how do I know that person," and I would draw another name and certain people just kept popping up, the same people eventually, and often was self-referential.

Some clients had introduced me to another client and so far the answer was sort of like, wow, the way I get clients is I have lunch with old clients and they have lunch with someone else the same week who has some kind of need, and because I just had lunch with them. I'm top of mind with them and they tell them, "You should talk to Konstantin." And that's how I got my clients. So the process was very repetitive. That's how it would work, but you can't actually target anything. It's like the circumstance of who do you happen to have lunch with and who do they happen to have lunch with.

It seemed kind of chaotic and so we kind of thought that People Map could bring some kind of structure to this mess. That was my part of the story, my inspiration first looking at the Six Degrees site that grew so quickly and then my own needs and seeing Friendster also grow so quickly. It's my motivations that came to the table. But we are five founders. Each of us from a different area.

So the next challenge was, Friendster already exists. They have millions of users and what if they just add a business profile to their site? That could be kind of a problem because network effects are very nice if you have them. They are not very nice if your competitors have them. And there was an established business network already. How many people know or remember RYZE? R-Y-Z-E? Okay. That wasn't huge by today's standards, but it was the business network that people knew at the time and they joined. So we had to figure out, one, since they didn't become huge, what did they do wrong that we should change, otherwise we would have the same, at best the same fate, as them. And then there was a company that got a bunch of funding called Spoke. Can anybody remember Spoke? So as they had a lot of funding, they're already out of the gate while we're still talking about who's going to be on the team and what we're doing.

So that kind of created a sense of urgency, though, to actually make it happen. And Reid had just left PayPal, so that was good, and he had some money. That's always good if you don't need to look for investors, and the founder can sort of put in some money, that just simplifies things a lot. But then there are all kinds of other companies that were already established. A company called Ecadamy, and there was Plaxo, which was doing not exactly the same thing, but kind of related, and Sequoia was backing them.

There was a company on the east coast called Visible Path, which was making a lot of noise, and there was a company in L.A. called Zero Degrees also making a lot of noise. So it was a lot of things, so in terms of how did we succeed? I don't think I have an answer for that other than I know why everyone else failed. So I think part of it is just, when you do a startup, try avoiding mistakes. And then, that may be enough because if your competitors make mistakes, then...so I don't have time to discuss each of the mistakes, but if anybody's curious about any of these particular companies, and why they didn't work, I think I've got a pretty good answer for that. And I think we, you know, we sort of succeeded by default because they all kind of messed up somewhere.

There was also a company called Jobster, if anybody remembers that. I know Minal knows them. Minal, here, and [inaudible @11:33].

Minal: I competed with them.

Konstantin: Yeah, I know. So and they said, gosh, you guys, and I was the VP of marketing and if you know marketing, you have to pick your target segment and just speak to that and have a very crisp value proposition. We always struggled with that at LinkedIn, because we're trying to connect the world's professionals for whatever they want to do. That is not a nice, crisp, message, and "so why should I use it?"

So Jobster kind of came out with this thing, we know what people, initially a lot of people who gravitated to LinkedIn when they're doing a job search and so they said, okay, let's just focus on that. But again, big mistake on what they did, I mean it bought them something, but it really didn't work. There was a company that never launched which was supposed to be huge and had a lot of funding, they bought the, they had the rights to the word Rolodex, I think it was Rolodex.com or something like that. We were kind of scared of them, too, but they, I don't think they ever launched. So all kinds of interesting things.

By the way, anybody know the original named for LinkedIn that we wanted to give it, but we basically didn't want to spend more than $5,000 for the domain, and so if the domain was taken, we had to ask, and if we couldn't buy it for $5,000 we just crossed it off our list. But the one we really wanted was called Well Connected. That's the name, but some networking equipment dealer had it, and Reid had the scheme of pretending we're trying to build some kind of content site to make it look like we were really poor, and he'd feel sorry for us. But he kept it. I don't know if he still has it. I haven't looked in eight years or so.

But that was sort of the initial name that we liked. There were a couple of other names, but in the end, LinkedIn was one of those compromises none of us really liked, but it was, nobody was so offended that they would hold it up. And then, over time, you know, I think people like it now and we always like the "in" part, we never liked the "linked" part too much. But "in" just wasn't available. So that's well connected.

Audience Member: Just "In"?

Konstantin: Yeah. Well, "in," we just thought has the emotion, that's why we put the little blue thing around it. We all liked "in" So then we're launching, and actually, one thing I'm quite proud of is, it's very hard to launch a social network in sales. But we're very proud that we were able to do that, because when we launched, nobody had heard of LinkedIn before. It was just all of a sudden it was there. It was great. It gave us a nice little pop in the beginning.

Reid knew Joe Ito and a lot of the bloggers, but the pop was a little different than we expected because everybody hated it, and said, "this is the most ridiculous thing I've ever seen." The bloggers are like, "Okay, they'll have users enough, and after a week, everything is going to die down." And they hated it because bloggers are very . . . they don't distinguish between personal and business. They're very public about everything, and LinkedIn was all about privacy, and controlling the contact, and separating your private life from your professional life, which was everything that bloggers didn't believe. So they, very quickly, told the world that LinkedIn would never work, which we didn't really care that they did. But it was like, please don't tell the world. Just keep your opinion to yourself.

But we think that there are more normal business people that you guys aren't. We built it for them, not for you can't argue with bloggers. It just makes things worse. So that was a challenge. Then we tried to tell the media a great story of what we had. Friendster was, it's hard to imagine but a lot of you remember what a hot thing Friendster w as in the Valley, and they're like, "How could you ever do this because sure teens and tweens friend each other but business people aren't going to friend each other on some site like this with photos. It's way too cute. That'll never happen. If it does happen Friendster is simply going to add a business profile and you're dead. So tell me again why I should write about you and what about your business model?"

I was like, "Well we're going to have a free version and a paid version and people will pay for that." And they're like, "The Wall Street Journal tried that and gave that up so good luck with that one too." It was quite hard getting the thing launched and the PR of that...actually we launched more when we got the funding and basically all the reporters who'd been talking about how great the internet was during the dot coms days all got laid off so the only ones left were the skeptics who didn't believe in the dot come stuff anyway. It was not fun. After that we decided he didn't want to be involved in PR anymore so I had the field to myself so it was great fun.

Audience Member: [inaudible 16:30] LinkedIn have on the other first round companies?

Konstantin: One cool thing is Sequoia actually approached us through LinkedIn and requested to give us funding. That's a nice thing. I think I would give that advice to every VC. Either people come to you and that's always bad, but usually the VC goes and searches out some deals and those are the ones that I usually end up doing, not when people come to them. As an entrepreneur you need to make sure someone tells them that they need to come to you so it doesn't look like you're actually trying to get money from them. I think Reid probably did some things so they would come to him.

I think the smart thing Sequoia did was use the product. Any entrepreneur is just so fascinated with his own product that if the VC actually uses it that just gives them a big leg up over other VCs. That's kind of important. I remember at Jaxtr, which was the company I went to afterwards, I was at this conference and a lot of people wanted to invest in Jaxtr and one guy pulled me aside. "Look here." Jaxtr was about clicking on the web and making a phone call, and he showed me his vacation rental cottage in Tahoe and it had a big orange Jaxtr button on it. "This is so cool. People are going to call me and I don't need to give out my phone number in case there's a problem." This is the guy I want to have on my board. For any VCs just try to flatter entrepreneurs by using their product. Sequoia did that well.

Audience Member: [So was that an intense training style? 18:01]

Konstantin: When they approached us it was probably 10, 20 and some other deal just took longer. They agreed after one meeting with Reid and then there was a partner meeting and a few hours later they decided to invest. But then it always takes two or three months of legal whatever. Officially we didn't get the funding till October which was 40,000, but really when they decided it was 15 to 20,000 around that time. But Ryze had 40,000 when we launched so we had caught up with them but were definitely not in the lead position at that point and Ryze, as a brand, was much better established.

Audience Member: So was it mostly PR that you were doing that was acquiring some more users for you guys?

Konstantin: No, it was a mix. There was PR. We didn't really do much PR before. The first PR was Rafe Needleman and Catch of the Day. He wrote about LinkedIn. That was the very first PR. You remember your first one. You probably remember your first one [inaudible @19:00] it was exciting at the time. There were one or two more but really we saved it for the funding to actively proactively reach out. That gave us a nice boost but we always watched everyday from virality how many invitations had sent out and our acceptance rate and we watched those numbers pretty carefully. Even though they were small the important thing was do they get better over time as we add things to it. In absolute numbers press was a big thing then but press always stays at a certain level. Even if you get better press you can never keep up with the viral loop. We could never get to 75 million users. That's that.

Design mistakes was promised, so I would say the two biggest problems of how we designed the product, we over learned. We over compensated. We looked at some competitors, Ryze was a very open system and everybody could send a message to everybody which a lot of VCs hated because now they got personalized spam. So we said, okay, we've got to put all these privacy controls in place so when you search for somebody we wouldn't even show you are they two or three or four or five degrees away. We just said, now I could introduce you either to this person or to maybe somebody who knows that person, but we won't tell you. Because we don't want Minal's contacts to be exposed so that I know who Minal knows. Because if we said, he's a second degree contact then I would know that this relationship existed but he may want to keep it private.

So went kind of overboard with that. And it was very good from a privacy and control perspective but the user's like, that's useless, maybe he's ten degrees away and he's going to pass it to nine different people. I don't want that. So we very quickly showed actually what the degrees of separation was. So that's one example. But I think we reacted very quickly. I think that was a good thing.

And the other one is we had a search feature right away and it was second day people said, why can't I search for people by name? And we're like, why would you want to search for a person by name? You know them, you don't need to search for them and they're probably not on here anyway. But cause we always thought like people want to search for meeting a certain kind of person with a certain title, certain industry and so forth. But the reality was a lot of people were just trying to. And we thought, like, well if you want your contacts to be on them, just invite them.

But a lot of people just wanted to search for their own contacts to check if they're already on LinkedIn. They didn't really want to invite anybody to LinkedIn, but if one of their connections were on there they kind of wanted to kind of link with them. And so we enabled the search. But then we were like, we should probably do like an address book upload because that's much, it's like an instant name search for whatever, 400 contacts that you have in your address book. And that's kind of how that feature came out.

So I think it's key to just learn very quickly what you missed. But not just to do what the user asked which would have been the name search. But to say like, okay, probably now we want to do the address book upload. And show you right away these are your contacts who are already on LinkedIn.

So those were kind of two design mistakes. I think the subscription model, I don't think it's that exciting when it's the usual advantage of recurring revenue is good when you have a startup. But I think the one thing that I feel is important, and it's probably not a good objective business decision, but I think as an entrepreneur I find it much more satisfying if users, if I know certain users pay for my product. I feel like I've really created something of value. If all I create something lots of people look at it and then advertise. Let's say I have some, I have a social network site and it's all free but Gap sells t- shirts on it. It basically means the users are . . . the old t-shirt, but the t-shirt that Gap sells is more about. Because Gap can figure out how to get money from these users, but I as an entrepreneur can't. It's kind of depressing, isn't it? That I have to go to Gap because they want Gap, they don't want me. I think subscription model is, or transaction model, whenever users pull out their credit card. It's just more satisfying as an entrepreneur.

The trick is, when you do it with individuals it's very hard. And individual's just don't really. And unless it's E-Commerce I don't really like pulling out the credit cards for it, especially for kind of recurring things. LinkedIn's genius from a revenue perspective is really the fact, if you go to enterprises you have a problem of all these gatekeepers and everything you've got to sell. And things get messed up in IT and purchasing and then people change positions. So that's really nasty too, the enterprise business.

But LinkedIn has got this beauty of where individuals adopt it, they use it for business, and then they expense it to the business. And so it's sort of an ideal thing. It's like, hey, I get these premium features but I don't have to pay for it. And I can use it both for business as well as to advance my own career. Now people like that. It's a little tricky with the companies. Do they want to spend for something that helps their employees find other jobs and leave the company? It's a little tricky and that was always something we had to navigate and we're still navigating today, so. But I think that's one of the things that made the paying model really work.

The future, basically people predicting the death of everything. That was kind of a fun article, it was ranked pretty high where someone took all the predictions of all the things that were dying, including the web. And but I'm going to predict the slow death at least, not the rapid death, but the slow death of the unsocial web. So I think when people look at social networking as one component of the web, I look at it as the web itself. And that all the other things, the rest, when I've been calling that social networking, I call this the web and this is the nonsocial web. And I think that will disappear over time because everything will be integrated. Because as humans we're the social beings and you can, that's already pretty far along with photos and videos and with resumes. But I think all the other aspects. In particular, if you look at all the top 50 service I think they'll all be connected with social information. Whether it's on their own side and they draw the social information in or they are little apps inside of the social network, I think that it will be smaller and smaller then on social web. And that's why the API success and that's why LinkedIn is best in that area, so.

The final thing about my own startup, it's still pretty early, it's a little more of a hobby so. I really mostly want to talk to people who actually want to work for the startup. At this point for free, no not for free, for stock options! Those can be worth something. Whether they are or not you never know. But, so I'm looking for people there, don't want to recruit anyone here from ZURB away, but if you have friends. If you have any friends. Someone's looking at, anybody has friends, you know, who are interested, I can share some information. And at some point, you can email me and once we're ready to let people in to test it.

We want to basically fly under the radar because I figure, LinkedIn was unusual. Because actually like what we started with is still what we are seven, eight years later. But that's really rare. Usually you take some huge turns especially in the first 90 days to a year and so I want to kind of stay as low key as possible so that we don't make all of our mistakes in the wide open. But with sort of 50 users at a time, so that's kind of a bit behind. It's not that I think someone's going to steal the idea or anything like that. That's the thinking there.

So, if anybody's interested, wants to try out things, send me an e-mail at, I don't know what it's, yeah, I guess Konstantin@Stanfordalumni.org. And if you know anybody who wants to work in some kind of crazy little startup, let me know too. And that's it.

Moderator : You guys have questions?

Audience Member: Well, once you got the 40,000 users, you got money from Sequoia, what were some of the key marketing things that you did to really start to grow?

Konstantin: Yeah, very untraditional. And, but it was kind of nice. Because I've been VP of Marketing at a couple of other companies and it's kind of cool because you get a lot of the money that investors give you and you get to play with it, a lot of people don't. But you're also the guy then at the board who has to say, hey, you're the one who sent all my money. So that's kind of the positive and negative side. So I basically didn't. I mean, we spent, you know the only thing we spent is we paid a PR agency $5000 a month and that's the only marketing spend we did for the first three, four years.

Audience Member: It was viral?

Konstantin: It was viral. So I basically worked on all the default templates, of what's the text when you invite somebody and tracking all that. In a fairly sophisticated way. We did e- mail marketing to existing members, the press, but it was kind of simple from that perspective. But also, till I left I was the only marketing employee, right? So four years I was the only marketing employee. Which is kind of fun. You get to kind of do everything and see everything, and, but you don't spend much money so.

Audience Member: Well, you got phenomenal press coverage.

Audience Member: How much would you say technology played a role? Because obviously there's creative stuff you had to do to really scale whole operation?

Konstantin: Yeah, I don't think the technology. I mean, it's got to work. I think that's the key thing. So that's obviously one of the challenges that Friendster has. Friendster's become this poster child of having so much demand and not being able to fix it, so. I think it's okay if you've got too much demand and your site gets really slow. But the thing is you've got to fix it within a certain number of weeks or months.

And at Jaxtr we had that. Our registration page took 30 seconds to load and we still got 30, 40,000 signups per day. So it was just kind of crazy. But we had to really work on the engineering side. It was never architected to scale that quickly, with the viral growth just overcame us. And we ended up hiring the CTO of MySQL to help us figure out. But we would fix one problem and all the load went to the next bottleneck. And usually with stupid stuff that was done by contractors or other things. But basically we just worked through one thing.

Audience Member: [inaudible 29:29]

Konstantin: Well, these days you can blame it on Ruby, right. Because they do these explicit joins of tables that should never be joined and you don't ever find out that rails is doing that. So, it's never you right? But the thing is, like, it was so weird because you like keep working on it, you fix a problem and it just moves to another spot. So at some point after working on it for a month or two you just think, like, is this ever going to work? But then all of a sudden, one of the fixes is done and all of a sudden it's working. So, it's, you have to believe that. You just have to work each bottleneck, and cause, the problems. I think of technology as more being sort of something you need to overcome. You can't fail at it.

But it's not what drives people to sign up to the website and things like that, so. But you got to be good at fixing it. And I think in consumer business it's very important to have a good analytical framework so you can, because all the testing of the templates was just a lot of AB testing and be able to get that data out pretty quickly. That's a good thing to try to bake in as soon as possible, or even better use some third party products that would work for you. Unfortunately Google analytics usually isn't enough so that's where the problem starts. It's really not good analytic stuff that handles this case.

Audience Member: How much were you trained to actually get through...

Konstantin: Not a lot, and I don't know. Maybe it would have been good to get rid of some features. There are usually not too many champions. You have to have a real champion who wants to kill things. And I could see it sometimes working. But we always felt like, hey, nobody's using that particular thing so much, that's okay. But I don't know if there was a cost, would it work better if we killed some? There are definitely some features that weren't used as much as we thought and, so, but by the time you get 75 million users even a small percentage ends up being kind of meaningful usage.

As long as the features accumulate data that can be valuable to the whole enterprise, it's kind of a good thing. Like even if they don't get used a lot. The data becomes very important. And I think that's one of the key assets I think. It has this data and LinkedIn is I think getting pretty smart and sort of using this data, you know mixing, slicing and dicing it and pretty soon LinkedIn will expose more of that.

That really is one of the core values that professional will get from LinkedIn is LinkedIn spitting this data back out after processing it in a way that really helps people drive their business forward and in their career. So I'm pretty excited. I think that's one of the key opportunities and I think LinkedIn is pretty focused on it, it's got a good team on the analytic side. They're always hiring there too, so.

Audience Member: How early on did you integrate with Outlook and how difficult was that? It seems like a great understanding of your user base...

Konstantin: Yeah, yeah, no it was quite early. It was not the Outlook toolbar it was just an Active X control. But that we built after two months or something like that. So that worked out okay. We probably waited too long on the web mail. And in some of the things we did I think what happened is with the viral growth, it's pretty good. We're never great at viral growth. Not like Friendster or My Space or Facebook, but it was always very predictable. But I think for a while I think the company just kind of coasted on that. And I think now there's much more of initiative to really optimize.

A lot of things like, I think a lot of the text that I wrote and I left as a full time employee four years ago or three years ago, I don't know, but are still in there. But they have initiatives to touch those and optimize those. But sometimes when you have good success you don't you keep focusing on like the burning problems. And part of it was the technology it was very solid, very scalable, and very performant. But it didn't allow for very quick iteration, so that's been a big thing that LinkedIn has been working on. It's trying to sort of look like so call it, changing the engine while you're gaining altitude on the airplane. It's kind of a difficult exercise. But, that's been taking a fair bit of time. But I think the users are going to see more of the benefits in the next couple of years.

Audience Member: What was your biggest fight?

Konstantin: Fight? Like with another company? Or internal?

Audience Member: In the company. Critical points where . . .

Konstantin: Well, there was a big decision to be made so we raised our Series B from [inaudible 34:05] at a pretty good upstep in valuation. And Reid gets a lot of credit for it. He's very good at fundraising and working investors, and he's obviously an investor himself. So he did well there. But we basically had no revenue and we realized that the next round we've got to show good revenue or otherwise, it's going to be a down round. And we already had a pretty decent burn rate and so we, it was kind of like under the gun to produce revenue and we sort of consciously decided to start with job listings.

We actually thought it was the worst revenue stream but at that time we had just gotten the funding so the water wasn't here yet. It was just kind of here and said, okay, the job listing thing is going to take a while. Like job market places take a long time to develop. And so let's start with it now. It's not going to get immediate rewards. The premium thing is going to give the immediate rewards. But that gives it time for the marketplace to build up so later it became a more meaningful revenue stream. Bit it means we really put on the pressure for the premium offering. And so, there are a lot of discussions about what premium model to use. It's actually Harvard Business School's case study on this particular decision.

So, we have a good section of the user base that we call relationship managers, and they do three things primarily. They reconnect with their old colleagues. They stay in touch with their colleagues, and they have a place to park their resume, and that's pretty much it.

A lot of people are saying, "Yeah, that's kind of me". And that's just what we used to do that and LinkedIn and keep trying to find out what else can we do to, at least, give them a fourth thing to do because it is kind of a passive thing. But then, there's two active segments; one, we call the searchers and one, we call the networkers.

The searchers basically, they can be recruiters. They can be financial investors doing due diligence. They could be people doing market or technology research or other things like that or competitive research. To them, LinkedIn is a people's search engine for professionals.

In fact, at Jaxtr, the founder, I always knew he was like a big LinkedIn fan. He contacted me through LinkedIn, so he is also responsible for me leaving LinkedIn. He sent a contact request through a mutual introduction, and that's how I ended up. He didn't contact me about a job. He was smarter than that. He just asked me for advice and then pulled me in the slippery slope.

I always knew he was a big fan of LinkedIn. He told everybody to do it. After a year, I told him, "Man, did you see the new thing we have on the home page?" He was like, "I've never seen the home page." So I was like, "How could you not have seen the home page?" But he basically bookmarked the search form, and he just goes directly to the search form, never looks at the home page, doesn't see anything that's valuable there, but he uses it like for Google for people. So, there is that segment.

Audience Member: Going back to the series B, you're into it.

Konstantin: So, that's one approach to make money off of this user base that are searchers. There's a much smaller group of them, like a couple percentage, but they would pay more per user. The thing was it would violate some of the privacy things that we had. Some people might react negatively, too because before you could only contact people on an introduction. There wasn't this ability to e-mail somebody that you didn't come through an introduction, which was poor core values. So, that was the down side.

The downside of going with the searchers was that there are fewer of them. There's no similar site that has proven that this model works, and users might be up in arms. At the same time, LinkedIn network used to be four degrees. We also collapsed that to three degrees to make this work. We basically killed two-thirds of people's networks so that wasn't really a positive aspect of it, revenue opportunity either, and there was quite a fire storm at the time.

The other one was the networkers which is a bigger audience. They're much more proven that they'll pay, but they usually pay, of course, a smaller amount. Those were the two camps there and with the networkers it was . . . the searchers and the relationship managers would get along well because basically the searchers only paid us because we had the relationship managers.

Even though these people are very passive, it's actually quite valuable to LinkedIn. Sometimes, people will say, "Oh, if you use your password, they're not that important." Well, okay, it's actually very important, but the networkers and the relationship managers don't get along well on LinkedIn because the relationship managers love LinkedIn because it's a clean, quiet, well lit place and it's good. They think they're at the office, and so, good.

The networkers think they're at a networking event. And so, they keep asking the relationship managers, "Hey, let's connect" and they're like, "Who are you?" Like, "No, putz, we're networking here. We're on LinkedIn. Go, get a life; LinkedIn. I'll go to Facebook."

So, they don't get along, these two groups because . . . the networkers call it the most boring social network ever because nobody responds, and they're not all like them. Networkers get along with each other, but they don't get along with people there. So, we could have monetized either of these groups, and that's was the case was. There's disadvantages of both, so that was a pretty hefty thing.

I don't think it was like any kind of fight where I felt like people did get upset and fighting with the other person. There was much more trying to convince and get the positive and negative. So, I think it was fairly constructive. I don't remember it being such a bad thing.

Moderator : Thank you.