Andy Carvin's Waste of Bandwidth

Snider: They see news as factual; people aren't supposed to comment on news.

Smith: It's like people weren't seen as sophisticated enough to comment on news, and that's just wrong. But it's definitely a conversation. The people who contribute the most are our top two percent of users. They're important, but we can't totally cater to them either since other people buy our paper and don't participate. So it's a tough balance.

Tatum: The editors always think they know what the news trends are and tell the audience. USA Today is now letting the user start to influence that voice. Do we need to find a way that users have their own areas to broadcast what's relevant to them, and should this area be separate from traditional news content?

Smith: Print mags are always being influenced by what's being said by the public. We take that info and find the best way to present it, explain it to a mass market. The benefits of comments system is you get more input in the editorial process. I think your process is always better the more you involve the public in the process, since they can help you sort out what's going on.

Tatum: What about news sites that are done entirely by their users? Do they need editorial staff or is it irrelevant?

Snider: I build user sites because we need user input. My site has two guys running it basically. We rely on the public to submit data, almost like wikipedia. It's moderated by people who come to the site and earn status. It's good to have editorial in the sense of giving people direction on how to participate. In the future, news websites will be primarily user submitted, because it's so cost-efficient. Three people can moderate 20,000 others. We have sites with thousands of submissions with literally one person overseeing the process. We let stuff in that may not be perfect, but the community will fix it themselves. We don't have to be grammar police. It's better to have a page with bad content than no content, because the community will work to make it good content.

Smith: People want it to be an easy process.

Snider: People want to do things their way. They have strong opinions.

Rafer: But this is stereotyping. Everyone has a topic that they're a freak about, but they want to be spoon-fed everything else. Some people know a topic, like chronic pain or whatever, but they don't want to participate in other topics. What happens when something gets fact checked two weeks later?

Snider: That's an issue with wikipedia. When was this edited and should you trust them?

Tatum: Sometimes there's no tone, no fact checking, and that can be scary for some of us.

Rafer: But many of these sites do have a tone. The community imposes the tone, publicly, organically, on their own timeline.

Snider: Editors on our site are just glorified users. The majority of edits are being done by three dozen guys, and among them, three of them just crush everyone else, writing guidelines, submitting every day. Because they're creating the site they always wanted to build but lacked the tools to do it. So we build the tools for them to do that. The secret sauce here in monopoly money. People come here for hours and hours, and if you don't give them a perk, they leave. That perk might be talking to us directly to give us ideas. That kind of rapport creates long-term loyalty. Other sites like Wikipedia have people working just to earn virtual gold stars - and they love it.

Tatum: What about actual compensation?

Snider: That's a big fear. Some day someone may actually demand money for all the stuff they did if the company gets sold. They want a piece of it. And that might change the rules of the game. The majority of people submitting content tend to be young, and that can be scary for some. Explains the bad grammar, too.

Davidson: So you want users to generate content but you fear them too.

Snider: Sure. But some of these folks get so good at what they do, we end up hiring them. They know our sites better than we do.

Tatum: Will we have ownership issues?

Davidson: Webshots.com had a large community generating page views for user-submitted photos. We used to love you could type in any location and you'd find a picture of it. The terms of service were such that you relinquished the copyright by uploading it. (Really bad idea -ac) At some point it'll be legally challenged. (As it should. -ac) Someone somewhere will argue that they didn't understand the terms and sue a big site (read: Boston.com's photo policy -ac) for these practices. YouTube is the biggest example of this. (Except that YouTube doesn't claim ownership - just commercial licensing rights. -ac)

Tatum: Will there be professional twitterers?

Williams: Well, we didn't think people like Scoble would take over like he did. Who knows.

Tatum: When you founded blogger, you didn't anticipate the commercial apps. Do media companies not need to be using these tools? Do they have to integrate or just use third-party apps?

Williams: News sites probably need to use more social media apps because it's in their financial interest.

Tatum: Why was Yahoo interested in mybloglog?

Rafer: They're pretty good at figuring out what people care about, except in the blogging world. We had sites that a couple hundred million folks looked at. We don't have much personal information of these folks, but we could make some general assumptions based on our data. Bloggers want to have a presence. They use our widget to create additional human connections. Once people's faces start showing up, it feels a little less lonely. We don't use a timestamp because new bloggers may only have one or two visitors. So we don't want to discourage them. About 70% who create a profile on mybloglog have a blog already. In a lot of cases, they have profiles on a bunch of social networks. But then there are middle aged British politicians or pastors whom we suspect just created their first social networking profile.

Tatum: For some of our users, like on Chowhound, that's their only community, and they scream whenever we make a change. That sense of ownership can lead to some friction. What are the best ways to pool those voices in without losing your own voice?

Snider: Well, you separate the brands, right?

Tatum: But we want them to have ownership, where they feel connected. What we do in separate branding sometimes makes it difficult to program for them because they think it's their site, not ours.

Rafer: Well, you treat the editorial folks as peers of the community, not taskmasters. (Not the word he used, but that was the idea.)

Smith: Paid editors can write reasonably well and the users can't, but they may actually have the same level of expertise on a given subject. You need to actively engage users and make them feel like they're a part of the process, give them that monopoly money, include their submissions in high-profile places so they feel they're rewarded for good input.

Rafer: Ignore for a minute who's good at it and who isn't. If users get obnoxious, why should you care?

Tatum: because marketers care. They need to feel there's a safe place to advertise.

Snider: Bulletin boards don't get much advertising but they attract tons of traffic. That's an obstacle. You need advertisers comfortable with that chaos. We're trying to build a resource where people are building topical resources, rather than let people do anything they want, like MySpace. We moderate to make sure they don't harm us; that's the threshold. Anything else is fair game.

Smith: It works for your site - comics. Does it work for other topics?

Snider: Sure. Gardening, pets - look at dogster.com. (Woof woofs from the crowd.)

Smith: What if you do a baby raising site and people give bad advice? Or pet care advice?

Dogster founder: People want to share opinions and they know they're just opinions, not professional advice. People have strong opinions about pets, like should dogs eat processed or raw food. After a while, you just have to say it's been discussed to death - agree to disagree.

Tatum: But you've gotten tons of advertisers. How?

Dogster guy: The circle of trust. It sounds goofy, but we bring in advertisers, help them figure out the messages they want to share, and bring it into the community in a way they would want. Not banner ads - things like contests and the like. Things the community appreciates.

Tatum: You can sell ads, sponsorship or advertorials. Top Chef came to us and said we want to be a part of your blog launch. They didn't seem to understand we can't just cover them as news and get paid for it. Otherwise our users will think we're total whores. You need integrity and trust. So we test strategies. Our community is adverse to marketing messages, so we need to be careful what will resonate. With Top Chef, we said we'd like to do a podcast or something unique. So it gave them an advertorial placement and it was useful to the community. AmEx did a campaign with us, interviewing chefs. Should be a perfect fit. But none of our users cared. They didn't associate that content as being relevant. So it didn't work.

Smith: Advertorial works when a discerning user can actually use it. In magazines, they get shoe horned in and no one cares. On the web, they can ignore it or use it without feeling like we're interfering with the experience.

Snider: Have you ever faked user accounts for promotional purposes?

Smith: I used to a long time ago but it was difficult to maintain.

Snider: Well, it's a form of espionage to see what's going on in the community. There may be some crazy users, but they use your site to be crazy, and that generates more traffic. It's kinda funny, actually. You ban them, then email every director in the company and say "fire this guy!"

Smith: But if you did something to generate that, then you're engaging them, and that's better than ignoring them.

Tatum: How will Twitter make money?

Williams: We don't. We're a tool provider rather than a content provider. With blogger, people used us for their own content purposes, they can pay us for more features, otherwise it's free. We don't care about editorial. They can worry about it. We have millions of blogger sites - we can't care about editorial.

Tatum: Does gmail freak you out when you get an ad based on what you write in your email? It creeps me out.

Davidson: Gmail is a bit creepy that way, but that's the nature of contextual advertising, and that's what they do. That's why we don't use google ads - it's just a little too creepy for user-generated content. We collect very little personal information. If we take your zip code, we can offer you feeds relevant to your zip, and at least that's useful. It's not scraping really personal info, like from your calendar or email.

Tatum: I went from reading certain sites to using my rss reader and aggregators. Digg is interesting because you get the story and the discussion. Is that a threat to media sites?

Williams: They're filters, pointing back to your site, driving your traffic. You may not go to your site directly any more.

Smith: but if you create good content, then the aggregators cover you and drive lots of traffic.

Snider: But they're also just popularity contests. Good content may not rise to the top.

Tatum: It rewards interesting content. But how do you know you can trust the content that's getting all the attention?

Smith: Well, if the comments on Digg say the story is full of shit, then you get a sense of whether it's good content.

Williams: "Professional" content isn't always factually correct either.

(quick break to Twitter and check email.)

Audience question: How effective are filters like Digg? Do they truly bring the best stuff to the top?

Rafer: They're subjective. Rafer doesn't work to me in the slightest; I don't go to the site. Others love, it, though. There are as many filtering systems as consumption habits, and it's all going to get spread out.

Smith: I spend a lot of time trolling these sites to see if there's stuff we should cover.

PublicSquare.com creator: What about reputation systems for users?

Snider: We use them as a way of filtering content. You get to a certain level by participating at a high level, and you get more trust because of it. It helps a lot. If we had to deal with the submissions ourselves without reputation systems, we wouldn't exist.