Before I copy and paste my liveblog from Notepad, I'm going to say that it rocked to meet Garrett and Josh because I admire them so. That is all.
Now, here we go:
Keynote Q&A: Joshua Schachter of del.icio.us & Garrett Camp of StumbleUpon StumbleUpon founder Garrett Camp and del.icio.us founder Joshua Schachter head up two of the most important social bookmarking and discovery services on the web. In this session, the two will talk about trends in social media, where the future may be heading and entertain questions from attendees. Moderator: Danny Sullivan, Editor-in-Chief, Search Engine Land
Speakers: Garrett Camp, Founder & Chief Product Officer, StumbleUpon Joshua Schachter, Founder, del.icio.us
These are two of my favorite social sites. I have been super-excited for this keynote. Now we're going to hear from Garrett Camp of StumbleUpon, a service that rocks, has over 3 million users, and was acquired by eBay for $75 million.
Garrett introduces SU. He says that he started it when he was in school but then got funding. He says that SU is not about search but about getting information based on friends' recommendations. There's such an explosion of content so you need filters to get through the extreme content to bring the best content to you. SU wants to create something that learns as you go to give you the best experience.
Joshua Schachter started del.icio.us in 2003 as a hobby. He is going to introduce his site as well. Key elements of delicious: tagging (social value of labeling). Delicious, has about 3 million users.
Joshua talks about how he created delicious in 2003 but it stemmed from a single-user creation he started in 2001. It's very hard to find things that you've seen before. You can go to a search engine and type it in, but delicious wants to make it easier for you based on other people's bookmarks. It's an interesting phenomenon because the attention of users is crystallized: people can recover what they've been looking for. A lot of people use delicious for personal memory but they also want to access what people they know have saved. Very little of the site is generic, unfiltered, 0-context social connections. However, the network pages (social connections) helps you see what other people bookmark where you can bookmark the same thing - you can see what's important to other people. It's an interesting way for people to have memory. It doesn't have to be you that recovers the item; it can be what other people have found. In that way, it's a social memory platform.
Now we're going to have some Q&A:
Danny: How would you define the line between social media marketing and spamming? Josh: A bookmark means "I pay attention to this" but it would look a lot different than something that is "Please pay attention to this." The community is less tolerant to that kind of behavior. It's not here for marketing; it's here for people to remember things. Garrett: We don't mind people marketing on SU but we have created a system to do that. Generally, if you have good content, you don't have to do much work. If it's not that good, people give a thumbs down, and it stops the story from being further promoted.
Danny: Both of the services have been sold to larger companies. How do you maintain the smallness? Josh: We have 3 million users so it's a big community to navigate. We want to make sure people have a pleasant interaction with the system. I don't think I could have gone this far without Yahoo. The infrastructure is far beyond what we could have built as a startup. To a significant extent, it is good pairing. Garrett: It's a relatively new thing. We try to keep the same team and we've maintained the same atmosphere in the office and things have been really good. When we get bigger, we will try to integrate additional resources.
Danny: What kind of role do you think these services will play in search? Josh: There's a lot of opportunity. Search has been in stasis for awhile. There's a lot of opportunity for exploring in dimensions and that social metadata is an important part of it. There's a lot of territory to be covered and a dozen things we can do from here. The partnership with Yahoo will help. Garrett: There's a personal social relevance. We want to shift from quantity to quality.
Now the audience polls the guys. I'm sad that they didn't ask my question.
Q: Based on recent acquisitions, your companies have changed - what are the positives and negatives? Garrett: Not much has changed besides finance. Productwise, it's exactly the same. Josh: We have a lot of technology that has helped us scale. The current engineering team that is working on delicious built Yahoo! Photos - this is an important consideration because we needed other access to resources so that we could scale for a large userbase. It's been a huge plus. In terms of direction, they bought us because they liked where we were going and haven't changed it.
Q: For Garrett - can you describe your first breakthrough? What pushed you out of this tiny idea and put you in the spotlight? Garrett: I think it was when we moved to San Francisco. When I was in Calgary, I wasn't going to networking events and I had about 500 users. Firefox is also a contributor - we were on the top 10 list and people started downloading it. There is no one event that drove our success. Josh: This is about hours and hours of effort for years and years. Delicious picks up more traffic in a day than it did in the first year. Followup: How did you know you were on the right track? Garrett: I really just built it for myself. But apparently it worked. Josh: I did it for myself too and it worked from there. You want tight feedback from the users. When you're the engineer and the QA, you learn very quickly. If you're using it, you need this tool and the turnaround time is faster. Followup: Is it hard to implement new changes now that you've been acquired? Josh: The sense of it is still intact. We know where it should go. Engineering does slow down a bit however. The new delicious has been worked on for a long time - it's the same delicious I built with more patches and layers. Garrett: As you get bigger, it will be a bit tougher because people need to get used to the new features. You can't just change something without an announcement. Get feedback from the users (take note, Kevin Rose!) before implementing any changes.
Q: If you believe in the value of the collective voting, why nofollow links? Josh: It's too much of a spam target otherwise, but that's it. It's not collective voting when someone registers 1,000 accounts and votes. There's no easy way to enforce that. We're not about building links, so we don't care. Followup: Would you algorithmically erase it? Josh: If I can prove that you're a real person, maybe. But we see it being gamed too often that it's tough.
Q: What's been the biggest "a-ha" for you when you rolled this out? How did they change the evolution of products? Josh: Delicious is very unconstrained. You can tag things the way you like. There are some people who tag things "to read" and then will access it later. People build workflow around links. People have also built chat systems around these bookmarks. Garrett: Some people have installed StumbleUpon extreme and they hack their profiles, but we don't have an API (awww - get one please!). In the future, we intend to (yay!) Josh: The current posting interface for delicious was some random guy and he said "you should do this" and we didn't listen. He built it anyway and it was a success: we still use it.
Danny: Your emphasis is on the bookmarking but you also have social networks. Josh: For delicious, that's only there as necessary. If you add someone to your contacts, it may be because you want to follow them on similar interests. We're not so much about a generic social profile. Delicious 2 is the same site on a completely different platform. Garrett: We're halfway between Facebook and delicious. We are about social aspects but we also like the content discovery. We're not going to implement changes just because MySpace or Facebook did it. (Thank you, Garrett. You rock.)
Q: I tested the paid version of SU and it worked quite well. How far are you going to take the paid program? Garrett: Pretty far. Less than 10% of our traffic came from that system last year. We want to datamine the feedback to provide more consumer insight. We want to guide marketing efforts based on demographic interest (some cities respond better than others for particular content). We're going to track this and determine why this content arrives. Followup: Do you have any negative feedback toward the paid side? Garrett: Not really. The quality is pretty good. If the site is not good, we actually select sites based on the best matching ad for the user. If the advertiser has a better piece of content, they're going to be selected over yours. Followup: Is eBay pushing you to push paid content as well? Garrett: Nope, not really. We're doing pretty well right now.
Q from Neil Patel: From delicious, if you have a lot of friends, it hangs. Is that a database issue? Josh: It's a Javascript error. Neil: Can I use delicious again? Josh: No, you're way too popular. :)
Q from Brent Csutoras: Is there a way to sign up for the new delicious? Josh: There was a form but it's full right now. We're going to do roll it out to the public sometime this year.
Q from Brent Csutoras: For the paid submissions, if many people vote those stumbles up and the campaign ends, will SU still serve the page? Garrett: Yes, if people like that content, then you'll do just fine. It's a way to get your foot in the door but the votes (and quality of your content) will bring your content. You won't be paid for the views beyond that point.
Q from Brian Wallace of nowsourcing.com: Any chances that we're going to see any major changes to the SU homepage such as recent Stumblers and recently popular. Garrett: We're planning on doing that but we're really focused on the quality of presentation. Right now it's not our main focus. But we'll try to give you more of a personalized experience in the future because people have demand for it.
Q from Brian Wallace: Any plans for combining logins from eBay with OpenID? Garrett: Not at this time.
Q: Can you talk about the concerns about private bookmarks versus public bookmarks? Josh: When I launched delicious, we didn't have hidden bookmarks until we were acquired. People need to understand that the behavior around these systems. There's a little less privacy now that you're much bigger. You can get closer to people's brains in a very direct way. Delicious is not explicit; it's implicit. It's not publishing - it's below publishing. I had a brief round a week or two ago and I was bookmarking financial sites. People were asking me if I went back to finance. It's terrifying and useful at the same time. If you don't want your information on the Internet, don't store information on sites that are connected to the Internet. There's some loss of perceived privacy. Garrett: With information systems, there's no scarcity until you get bigger. You may want to share stuff just with family and friends. You may want to keep things private and other things public. Josh: People want things with a lot of features but they want it as simple as possible. Garrett: Can you build an interface that integrates both? Yes. But it's something we might add in the future.
Someone said something about how delicious is great for collaboration. He said that they discussed podcast informaion (which I've done as well) based on delicious's tagging. It really is useful, guys.
Q: Have you guys ever thought about unifying these networks in the future where you can have one profile that can be merged? Josh: There are a lot of things that people call social bookmarks that are not. I read Digg all the time but it's not social bookmarks. (I agree. It's social news, dudes.) I don't think that StumbleUpon is a competitor in the slightest. There are a bunch of things that are referred to the same way but they're not related. Garrett: There are different use-cases. I don't think that one company will be able to do this really well.
Danny: Do you view yourselves as rivals in the social media space? Garrett: We learn from each other design wise and interaction wise. We all share semantics while we figure out what makes the most sense.
