The cofounder of the web’s fourth-largest site tells what’s wrong with our notions of performance-based advertising and user-generated content.

It’s hard to say which is more surprising: that the fourth-largest website has only two dozen employees or that it doesn’t make any money. But one thing that almost never raises an eyebrow about is that it’s outrageously popular.

In eight years, the site has become the ultimate example of user-generated content with 11 million articles written by 100,000 volunteers – strangers speaking 250 languages who pledged to adopt a neutral point of view as they work collaboratively to document the often-contentious world around us.


The U.N. should have that kind of support.

Co-founder Jimmy Wales is justifiably proud. But money is good, too. So he started the for-profit four years ago to create “the rest of the library,” and hopefully to turn a profit through contextual advertising.

Like its older sister, Wikia lives a Spartan lifestyle in a humble loft in the heart of San Francisco’s Multimedia Gulch. Desks are scattered around a large open room – no cubes, no dividers and only a few bland offi ces for private chats. The color scheme – avocado, eggplant and orange – might be more appropriate for a health food restaurant than a post-meltdown startup. With a few beanbag chairs and a foosball table, it could pass for a dot-com circa 1999.

But the simple scene masks complex thoughts swirling around inside Wales’ head on topics like performance marketing, social media and user interaction – areas where Wales has demonstrated a strikingly accurate sense of what lies beyond the next big thing.

You’ve expressed fears that online ads might go down the “rathole of direct marketing.” What do you see as the danger there?
That terminology comes from John Batelle of Federated Media. I don’t think he’d say he’s concerned it will, I think he’s convinced that it has, primarily because online advertising of various kinds is so measurable in a lot of ways. You can measure impressions, you can measure click-throughs, you can measure conversions. That has been a real draw for people to focus on that area because we can.

If you are selling billboard space beside the highway, you really have no good way of measuring how many impressions you get – although you can estimate that. You certainly have no way of estimating how many “click-throughs” in terms of actions taken to get more information. And beyond that, you certainly can’t test conversions – to say that because you had a billboard, you sold more BMWs. There are ways to measure things like that, but it’s traditional, old-fashioned thinking about brand awareness.

I like to use the example of Facebook. As everyone is aware, Facebook has had some difficulty in monetizing as opposed to, say, Google. One of the reasons for that is because when consumers are on Facebook, they’re not shopping at that moment. Certainly, direct response marketing can be very lucrative if you catch a customer at that very moment where they happen to be shopping. That’s a great moment to pitch them with something.

On the other hand, I think we’ve overlooked – and this follows from John Batelle’s thinking – that brand advertising actually matters and is really important. What’s interesting is that it’s really cheap online. If you want to get access to 20 million college students, you can do that a lot cheaper on Facebook than you can on television or other media.

You’ve used the example of Doritos for that. It’s one thing if you want to catch someone while they’re in the process of buying a laptop and say “Hey, look at HP.” But it’s another when you just want to talk to college students and get them to think about “Doritos.”

That’s right. What I would suggest is that for Doritos or similar brands, to get your name in front of people, or even engage in innovative campaigns, it’s a lot cheaper to do it online than on other media. And I think the reason is that the whole Internet industry has geared around this incredibly lucrative direct-response, keyword-based advertising dollar. It makes sense; Google goes after that because it’s good money. But I think there’s a whole other part of this that’s been overlooked.

It looks like Facebook is making inroads now by coming up with more strategically marketed campaigns with companies like Adidas, Papa John’s, or Coca-Cola, which has something like 3 million fans on the site now. Personally, I’ve never considered myself a “fan” of consumer products. What do you think about that?

Certainly there are people who are “fans” of all kinds of consumer products. They really do love them for some reason. The term I use, which I assume I didn’t come up with, is “information-dense” products. What that means is, if the product is one that consumers can talk about – that there’s a lot of information – then they’re much more likely to become engaged with it. I think Coca-Cola’s an interesting example as compared to Doritos in the sense that Coca-Cola has a certain sense of history, a lore, part of Americana. It has cultural impact – positive and negative – all over the world. I think there’s more symbolism in that brand than some others. So it’s fairly unique in that regard.

I think that lore comes out of very traditional marketing, from building that brand up for such a long time. Do you think that’s even possible for a product coming out of the gate today?

I think it’s a lot more possible than ever before, simply because it’s possible to engage with people in a really in-depth detailed way that was really hard to do with a 30-second commercial. You can actually tell people your story, and you can have a story that’s worth talking about. This is one of the things that I’d hit upon, that authenticity matters more than ever before. Some made-up puffery isn’t going to get you very far. But having an interesting story that people want to talk about is very valuable.

Why don’t more brands do that?
I feel, and I keep hearing, that a lot of brands are still afraid of the Internet. They’re afraid of advertising on user-generated content because they have a wrong model of what’s going on online. The model [they envision] is angry message boards with a bunch of trolls yelling at each other. Their model is low-quality Geocities pages with flashing blink tags and all that. That doesn’t really fit the reality of what we’re seeing today.

We’re seeing a lot of high-quality stuff come out in a lot of different places. There’s still crap on the Internet, but I think what we’re seeing is a lot of high-quality projects coming together. If you’re BMW and you’re thinking of advertising on a BMW message board, you may be a little hesitant because you’ll wonder “Am I going to be putting my ad next to people who are screaming at me?” But even message boards have gotten better, I think, with new moderation techniques and software techniques to monitor discussions.

I think everyone has been on an unmoderated mailing list that became unusable because a few very loud, very noisy people who loved to argue with each other came to dominate it. And the quieter, intelligent, very kind people drifted away. I can remember when it was really a struggle: how do you moderate that? But I think we’ve gotten better with message board software where people can be rated lower by other users and moderators have learned to moderate with a gentle touch.

Let’s talk about how the evolution of Wikipedia and Wikia tie into this discussion about getting people engaged.
Wikipedia is now a global brand that is known everywhere that people are on the Internet. It was built without spending any money at all on advertising – zero dollars and zero cents – essentially through word of mouth. More than anything else, [it was built] by having a story that is interesting to people, that is worth repeating and worth believing in. So the idea of a free encyclopedia for every single person on the planet has captivated people as something really worthwhile to do. That’s really been propelled by people telling other people, to explain it to them and get them involved.

We see the same thing for Wikia. It’s very community-driven. People are passionate about a particular thing, like the TV show Lost and “Lostpedia,” for example. They are super-engaged in this TV show because there’s a lot of stuff to talk about. And there’s a lot of stuff to talk about with other people who are involved in the TV show about “Lostpedia.” It’s a community project. We know absolutely that J.J. Abrams, who is the creator of Lost, is a big fan of “Lostpedia.” There’s a symbiotic relationship between the writers of the show and the wiki because the wiki documents the entire history in a way they don’t even have documented internally.

The fans keep track of everything about that universe and they’re really concerned about little inconsistencies. It’s the kind of thing where any discrepancy might be an error, or it might be a clue, because the whole show is quite mysterious. So the fans obsess over these details, and it matters to them that people are reading it. There’s a symbiotic relationship with the actual show, and it might even influence it, which is kind of cool. So it’s those kinds of things that tend to work very well as communities. It really is about having something worth talking about.

How granular can it get? You’re talking about a virtual community with Lost. Can it be a local community?

Yes and no. It’s a little slower to get started because there’s a critical mass issue. But what I’m finding is that where it works is where there is a dedicated group of people who are committed to something. For example, there’s an SFHomeless wiki, which is about homelessness in San Francisco. It’s a place where people who are involved in the local community, dealing with homelessness – church groups, shelters, government agencies – come together because very often people don’t know what the resources are that are out there. So if you’re a pastor in a local church and you find out about someone in trouble, you may not have a very good set of knowledge about where that person should go and what’s available. They just want to have a centralized database where they can hold all that information. They’re very passionate about what they’re doing; they have something in common.

So that’s working very well. What we haven’t seen working very well, so far, is a general wiki about San Francisco with all of the events and things like that. And that’s simply because there isn’t a community of users. It doesn’t click for them. I think it will at some point, but so far, reporting empirically, we haven’t seen much success in that area. I actually suspect for some of these kinds of citizen journalism initiatives, it may come from being even more hyper-local – not from San Francisco, but from my condo association, my neighborhood, where people actually begin to get together about things they can actually impact and that actually impact them.

Schools can be like that.
Schools are a good example. One thing that occurs to me – and this is just one man’s opinion – is that I never, ever, ever watch the local television news because I find it absolutely useless to me. It has no bearing on my life whatsoever. Even at the level of reporting on local politics, it’s both too big and too small for me to care about. It’s too big in the sense that I don’t really think I can have any impact on the local politics of, say, St. Petersburg, Florida. I’m not an activist, I don’t have time to be an activist, and they’re not really interested, so it’s too big. And it’s also too small. It’s not like global issues of what’s going on in Iraq where it seems sufficiently interesting for me to keep up with every day. And, of course, the local news is notorious because that kind of local political reporting is not nearly as big as the local car wreck and crime scene reporting, which is really useless.

The only time we care about crime is when it’s on our block, unless it’s some horrid crime that challenges our humanity. Then we care.

Yeah! And if it’s that interesting, I hear about it on the national news. In Florida where I live, there was the Terry Schiavo case which was local to me, but it was also national news. So most local news has no impact on my life at all. But near my apartment, they’ve changed the configuration on my road in a way that is extremely annoying to me. I want to complain, but there’s no place to complain. This is the challenge of a lot of these hyper-local sites: how do you get a critical mass? I have a feeling that a lot of people who live close to me are angry about it.

A lot of our readers are publishers who are trying to create content. They like user-generated content because it’s cheap. But there’s a qualitative difference. And Wikipedia, I think, is of surprisingly high quality. And I think it’s precisely for the reason you cited: that people are passionate about these subjects, they care about details, they don’t want to see errors entered into the item. But this is something that has eluded many affiliate marketers. How do you get people to start writing for you? How do you get people to help create content you can use?

It’s a challenge because with citizen journalism, there is an interesting chicken-and-egg problem, which is that I really get much more enthused about writing if I believe someone is going to read it. It makes a huge difference. I’m not writing just to amuse myself. I really want an audience to read what I have to say. The audience that may satisfy a writer may be different for different people and for different types of things. Maybe I’m happy if I have 100 readers for my blog if they’re people I admire and I’m writing about something I care about. But a lot of times, a lot of local journalism in particular is not that glamorous. It’s just grunt work.

Somebody’s got to cover what the city council’s doing. Usually it’s not that interesting. Occasionally you get a really good story out of it, but usually it’s just a debate about this or that. You may have an opinion, but it’s not that gripping and people don’t get that engaged with it. The only people who may get engaged with it could be people who have some serious agenda – and that’s problematic as well. I kind of like the idea of a fresh-out-of-school reporter being sent to do some boring stuff, mostly because they probably don’t have a strong opinion one way or the other. They didn’t self-select because they have a problem with the city council, and that’s the problem with citizen journalism. I don’t know what the solution is, but maybe it’s in a hybrid model.

You’ve made an interesting transition from the nonprofit world of Wikipedia to the for-profit world of Wikia, where I assume you now have your eye on maximizing profits. How does that change your view of the content and what your editing staff does in terms of massaging the content in the direction of the site?

Very little, actually. We basically follow the lead of the communities to write about whatever they want to write about. We view ourselves more as the provider of a platform than anything else. The way I look at it is to use the analogy of a bowling alley. Some people say, “Why do we do this for free? Some people get paid a lot of money to do this.” And I say, “Some people get paid a lot of money to bowl and some people actually pay to bowl.”

What are they paying for? What are they doing there? And what is the business of a bowling alley? The business of a bowling alley is not to crowd-source the work of bowling. If you open a bowling alley and you say “we’re trying to get some bowling done here” and “we’ll trick the general public into doing the bowling for free,” that’s crazy! And with user-generated content, that’s also crazy. You’re not trying to crowd-source the production of content. What you want to do is provide a facility that people enjoy. What you want to say is, “Look, we don’t really care that much what you’re doing, except we want to empower people who want to do certain types of good work. We want to make it fun and make it a good place to do some good work and to create something you’re really proud of.”

In a bowling alley, you’ve got to keep the lights on and the floor polished. And if someone gets a little too drunk on the beer, you’ve got to kick them out so that other patrons can enjoy their game. It’s the same with us. We work on the software to make it easier to use. We listen to the community, asking what do you need, what are the features you’re looking for? We help them block bad people. We control spam and abuse. It’s basically trying to create a place where people say “I feel happy. I’ve met some cool people. We’re building something interesting. You people run a cool service and it doesn’t cost me anything.”

That means certain things when we think about the content that’s created. We don’t control it. These people aren’t working for us for free, which makes no sense at all. We can’t make them do anything. On the other hand, just as a bowling alley might offer a pamphlet of bowling tips, or even bowling classes, we can offer some guidance and help. We can say, “Here’s our experience. Here are some things that work.” So when someone is starting a new wiki, and says “How do I start my wiki?” or “Someone is causing trouble. What do I do?” we can say, well, here’s what has worked in the past. It’s really more of a community-building business than anything else.

What do you do if you have a community that is sort of taking over? What if some sort of political extremists, for example, start ruining your wiki?

Well, we don’t. So it’s a hypothetical question. We have over 12,000 sub-communities. If political extremists came to the Muppet wiki, the Muppet community would just kick them out. We give our communities the tools they need to manage things. But also we would say, maybe you’re really in favor of Bosnian independence or whatever. You can make your own wiki for that.

For example, we have a wikiaGreen that has advocacy for sustainable living, for green technologies, for green ideas. Basically, that community has a certain sense of what’s appropriate and what is not appropriate. Part of the answer in that case is to say, “Look, that’s a great debate to have, but this is not the place to have it. Go have it somewhere else, go start your own wiki. What we’re doing here is advocating. And if you come in and you’re a dissenter, maybe there’s some ways we can work with you. But in general, you’re probably going to be more comfortable going somewhere else because this is a group of like-minded people trying to create something here.”

What about the blending of forums with the wiki model?

We do offer forums, but for the most part, people come to us because they’re tired of forums. They’re tired of just debates that never go anywhere. People are really more interested in building something together. At Wikipedia, people often meet others who, somewhere else, might have been encouraged by the software just to argue with each other. Now, instead, they’re encouraged to collaborate, to produce a document that everyone agrees to that is neutral. So if you come to that issue and you say, I don’t want to be propagandized to, I don’t want advocacy, I just want to know what’s up, the collaborative system really brings out people like that. If people just want to debate, there are lots of places to do that all over the Internet.

When we started MarketWatch 12 years ago, we found people were reluctant to come into communities, which surprised us. We thought people would want to discuss stocks and investment strategies. Now that’s changed, and it’s become quite popular. Have you noticed other ways that people have changed in the way they use new media?

Yes, definitely, especially as we take a look at something like Wikipedia. It’s now pervasive in a lot of ways. And people are increasingly educated about the strengths and weaknesses of Wikipedia in a way I think is reasonable. I think media reports about Wikipedia are better than they were two or three years ago, both in terms of praise and criticisms. I think they’re beginning to understand it for the first time. Increasingly, if I see a criticism in the press, I say, OK, this is a problem we have.

Another part is, I’m always seeing this question out there: Should students use Wikipedia? Should journalists use Wikipedia? I always say that’s the wrong question. “Should” implies they might or might not. They will. They are, for better or worse. The real question is, how should they use Wikipedia? What are the strengths? What are the weaknesses? How can we quickly communicate to people the understanding they need to get the best out of Wikipedia and avoid the worst? Those are the things that I think we’ve progressed on.

I think people forget that traditional encyclopedias also contained conflicting accounts of war situations, political situations, economic situations.

Absolutely. You get more than one thing. First you get simple error, which is far more frequent in traditional encyclopedias than people ever realized. That’s not a criticism. Britannica is a fantastic encyclopedia, but it’s full of errors. The best journalists know that even the best newspapers contain errors every day, despite very high quality.

When we think about experts, I always like to give the example of an encyclopedia article about the field of psychology written by Sigmund Freud or B.F. Skinner. We can acknowledge these are giants in their field. These are people who had a huge impact on their field with brilliant insights, but very much in their own, very different schools of thought. And, frankly, I don’t want either of them to write an encyclopedia article. An encyclopedia article written by either of them is a fabulous, wonderful, historical document – “This is what B.F. Skinner had to say about psychology” – but it’s very likely that would not be a contentious work.

How do you control that in a wiki? Today, most people are probably a bit more behaviorist, so how would you control that, when your editors and writers lean toward behaviorism?

Well, there’s no such thing as perfection, but you try to control it. Are you familiar with the movie 12 Angry Men? I think that, in its own way, is a model for a useful Wikipedia debate. If we’re doing our job well as a community, socially, where we really focus on quality and neutrality, where we really focus trying to really accommodate things, basically one person with thoughtful arguments, who can write well and cite good sources, can actually make sure an article is balanced, even against people who don’t agree.

And the people who don’t agree, if they’re good Wikipedians, should be able to say, “Yeah, we don’t like George Bush, but we do understand the entry about George Bush has to contain both criticism and praise. If critics are saying this, we have to include the administration’s response.” In my opinion it’s easier to do if you have an open system where you have rules that require people to take seriously any thoughtful dissent. There’s nothing magical about it. It happens because people care enough to uphold a set of rules about neutrality. Nothing’s ever perfect. We haven’t discovered the means to overcome human frailty.