Home Articles Transcript: interview with Scott Karp, CEO, Publish2

    Transcript: interview with Scott Karp, CEO, Publish2



    Interviewer: Paul Ryan, Editor, Anthill magazine

    Interviewee: Scott Karp, CEO, Publish2
    Interview conducted on 14 February, 2008

    Paul Ryan: In your blog post launching Publish2 back in August 2007, you wrote:

    Standing on Digg’s shoulders, Publish2 solves the WHO problem by creating a platform for networking the one group of people who are disproportionately more likely to be effective news filters across every conceivable topic: JOURNALISTS.

    Now that’s what warms the cockles of a journalist’s heart.

    Scott Karp: (laughs) That’s right.

    PR: Does it warm the cockles of everyone’s heart online? Have you encountered much opposition to the idea as perhaps being somewhat elitist?

    SK: No. The funny thing is, people talk the ideology a lot, I think more to push the boundaries. But most people, even the most seemingly dogmatic, are at the end of the day very practical. The perils have become apparent in a completely open, free-for-all system like Digg, which is just a constant battle against gaming from bad actors. It’s not about saying that X group of people are smarter, better, faster, stronger, whatever. When you give people an open system, they in fact don’t self-organise into something that is broadly useful and always benign. They self-organise around niche interests, sometimes in a parochial way. The law of the web is as soon as a system becomes worth gaming, as soon as there is something to be gained, people will flood in to do so.

    A first approach to syndicating our content is basically having people apply a specific tag to all the bookmarks they do and then we create a feed of that tag they can publish on their site. And someone asked me, ‘Can somebody else post to that tag, or go into my feed?’ And I said, ‘Well, right now, yes it could happen.’ We’re going to create secure feeds as we build out a newsroom platform. But the chances of someone doing so on purpose are highly unlikely because all you’ve got on the system are another bunch of journalists. The other nice thing is that everybody is in there under their own identity, so if somebody did add something to your feed you’d know who they were and we can ping them and say, ‘Hey, what’s up with that?’ So, you don’t have these anonymous accounts problem.

    Journalists are largely disenfranchised on the web. The way Google works and the way Techmeme works is they are all just reading the open links on the web and counting those links as votes and trying to figure out whose link should count more than other people’s links. The limitation of that system is that there are a lot of people with knowledge who don’t put links on the web. So, their knowledge is not being harnessed in any way. Publish2 is a way to give a platform for people to do that, who aren’t doing that.

    PR: I probably launched straight into the middle as I often do but I think you should go back and tell me about the germination of the idea. Did you wake up at two o’clock in the morning and sit bolt upright and think “Ah-ha”, or is this something you actually discussed over a number of weeks with colleagues.

    SK: The credit for the origins of this idea actually goes to my partner, Robert Young, who is the co-founder and the chairman of Publish2. He had been blogging at GigaOm since 2005. He was actually the first person that Om Malik brought on board, talking to the companies of various backgrounds in 2006. They still are trying to get their strategic bearings. He talked to one large newspaper company. Even back then the financial decline of the newspaper industry had begun but we hadn’t reached that level of acceleration, but there was still a great deal of concern about the overall direction and survival. And one of the online phenomenons at that point that was on everyone’s mind was Digg. So, wondering is this really a new viable editorial model? Is this going to replace editors entirely? Are the people just going to decide what they want to read? What do we do? Okay, so we get it. It’s a powerful model, it’s having a big influence on the web, but we do not know what to do about it other than stick this Digg button on our site. That’s where the seed of this idea was born.

    The big question on every news organisation’s mind was: how does journalism stay relevant? If people are going to spend their time with interesting stuff – the things that just amuse them – where’s the support forum for important journalism in that model? So we thought, why not just put journalists in the driver’s seat? So it’s a new networked editorial model. Robert contacted me at the very end of 2006 / early 2007. Much of what I had been writing about on Publishing2.0 went to where this whole idea was going. In many ways it’s fundamentally counter-intuitive to web 2.0 ideology, which is that everything has to be purely open. It’s counter-intuitive to traditional media companies and the dynamics that were based on monopoly distribution.

    So we started kicking around the idea in early 2007 and evolved it over time. By the middle of 2007 we decided to make a go of it. I did the scary thing that every entrepreneur faces doing at some point and quit my day job. So here we are.

    It is one of these concepts that, to me, is like a fine wine – it keeps getting better with age. It keeps getting better the more engagement we have from people out there. The whole focus is taken really on newsrooms as users. It’s a whole new dimension that has really evolved and it is really an exciting concept – not just networking journalists as individuals, which I think remains the core of it, but also networking newsrooms, networking editorial groups. Enabling them as individual editorial groups to do something that I think fundamentally every content brand has to do, which is not just publish its own content but help people find the best content wherever it may be. But to do that, not just in a traditional media silo. A lot of people have asked me recently why we don’t just create our own Reddit. My answer to that is that you basically end up with an old media silo. You end up with just a little island with no connection and no influence.

    A lot of different editorial groups are very focused on election coverage. Now they can cross-pollinate each other and help each other find interesting items to bring to their respective readers. There’s another big vision in here that looks at: how do you re-invent the distribution of local content? The wire services do not really make sense online because now you’ve got the same content syndicated and published on a thousand different websites, including Yahoo and Google. When there was only one newspaper in each geographic region that made perfect sense. It was a great model. Now what would make more sense is if all the attention to a particularly noteworthy local piece of local reporting can actually go to the site where it was published rather than having to be rewritten or redistributed somehow. That’s the real fundamental challenge for the web right now. It’s not publishing content – publishing content is now free. The challenge is distribution.

    What was happening with election was some of the newsrooms were bookmarking their own stuff and not putting them into their own feed, but just putting them into Publish2 so that other newsrooms could find it and add it to their feeds. You see the seeds of a model there. A piece of reporting in one locality gets national or international attention, it gets put into Publish2, other editorial groups discover it, put it out in their headline feeds and suddenly if you have a Digg-like traffic driving scale. You can be driving all that traffic back to that one piece of content where it was originally published – and you could be doing it in distributive fashion. The link to that is now appearing on dozens of different news websites.

    PR: I want to get back to the issue of filtering and control. If this really takes off and almost every article published by a news organisation or a blogger is registered on Publish2, doesn’t your filter go out the window? Doesn’t it just become a duplication of the news web, like, say, Google News?

    SK: That’s where the Digg-like effect comes in. Something only hits Digg’s front page if it gets Dugg by a certain number of people. Right now these headlines are basically just a stream of news – it’s unfiltered. But what can happen as the thing starts to scale up is that there will start to be overlap in stories that get bookmarked. We’re also going to introduce an overt voting feature where you simply vote on it because you want to get attention in a more purely Digg-like fashion. And our site already enables this – you can sort things by the number of votes or bookmarks at this point. That’s where you start to see the filtering and the prioritisation.

    So once the election effort gets up to scale, you will be able to rank it and see which items have been most frequently bookmarked, and then introduce the element of people actually going in and voting for those items. The dynamic of Digg is that when people realise the ability of the system to drive a lot of traffic to things that they think are newsworthy, then you start to vote on them because you want them to hit the front page. There are many different elements driving that ranking, from an individual journalist bookmarking something for their own reference, to a newsroom bookmarking something that they want to go on the headline feed, to any user of the system deciding to vote on this in pure editorial fashion, because it is worthy of attention. By now you have many different editors, but you don’t just have a free-for-all. The pendulum has stopped in the middle where on one extreme you have a pure open gameable free-for-all and on the other end you have a non-scalable, hierarchical command and control editorial decision-making process. We’re aiming for the middle ground.

    PR: Early on in the beta you are only letting “registered journalists” in?

    SK: Well we are basically doing it by hand right now. We’re taking two approaches. One is anybody who is working for a mainstream publication of some kind. And then bringing in independents. It could be freelancers, bloggers, anybody who is not employed full time. What we did with Knox News is one example where they brought in a group of bloggers, but they hand-selected them and brought them in under their banner. In a sense, taking editorial responsibility since they were actually going to publish that output. So that mainstream news brand took responsibility for these independents, trusting these people to act journalistically in their selection. That is how the thing can become more scalable and I think more appropriately porous. I think the essence of it is not that we ultimately get down to some endless battle about what constitutes a journalist, but more that we maintain the integrity of a defined community. Everyone who is in there is in there under some auspices or for some manageable reason. Ultimately, the way we really scale it is to give each individual user the ability to invite somebody else in, but since everybody is in there under their own identity, you have to take responsibility for anyone you vote in. If that person ends up being a bad actor, then that will reflect poorly on your reputation.

    PR: Or if someone starts out as a large news organisation journalist and then “goes native”, as they say, and becomes a fairly aggressive blogger, it may not have been your fault if you vouched for them because they used to work in the cubicle next to you and now they’re firing salvos at old media.

    SK: What Digg has shown is that the gaming of Digg doesn’t happen through any one individual bad actor. It happens through networks of people participating in a game. (eg: all friends, all Digging each other’s stuff.) So, one person can’t really do a lot of damage because it will ultimately take a large scale of votes for something to really get attention. Organised gaming will happen much less frequently than Digg has to deal with and I think we will be able to manage that. No system is perfect. Every system is going to have its flaws that need to be managed.

    PR: Digg seems to be quite prominent in your original planning and even now you’re thinking. Have you had any feedback from the people at Digg?

    SK: No. Silent.

    PR: You’ve not sought it?

    SK: To contact them?

    Just to see if they have had any thoughts on what you were doing and even criticism. You wouldn’t have to take them on board, but I would just be curious to hear what they think about the emergence of potentially rival variations on what they are doing.

    SK: Well, I think if they have any thoughts on it they will reach out.
    PR: They know where to find you.

    SK: They know where to find me. I’m not hiding what we are doing. Ultimately, I have all the praise in the world for Digg because they proved that this model can work and scale as an extremely powerful filter. I think Digg’s biggest problem right now is that it is stuck from a topic perspective. If you are a twenty-year-old male into technology, it is a fantastic filter. If you were still hoping that Ron Paul is going to be President, it is a great filter. But beyond some finite number of interests, it hasn’t been able to break out of that because its community is focused on those things. Digg is hierarchical and they know it and that is why they are trying to break the hierarchy to flatten it out and extend it more. They may well succeed. I think that trying to make an open model work is still a very worthy experiment and I think there is still a lot of learning to come out of it. I mean, they may succeed in extending into other topics. They are just fundamentally different approaches and fundamentally different ways of thinking about networked systems.

    PR: I suppose it depends on your sample size, but journalists have their particulars interest too. I mean they are probably more likely to write about national politics or culture or maybe technology than they are about, I don’t know, science. Across your most popular front page in twelve to eighteen months time, is there going to be a predominance of one, two, three subject matters?

    SK: I think our main front page is actually the least of my concerns because I think that people come online with particular interests. That’s why search is so powerful, because people know what they are interested in and seek out what is new on a particular topic. If the science articles don’t typically make the main front page, there are always many different ways to dig into science, from a top-level science topic to many different tags that will become popular and signal to people that there is interesting content on those topics. People who are interested in science will always be able to find science from the point where we get enough science journalists involved. I think that obviously, at some point in the future, for that main front page to be like a real newspaper, it does have to represent a wide diversity of topics and priorities and interests. But that is at the final end stage. One short-hand way we used to think about this is using something like a Techmeme, which became really obsessive place to go for people interested in technology. You do not go there expecting to find sports. There’s technology. There’s no Techmeme for science because there are not enough science bloggers to make that happen. So the potential for a Techmeme for science is actually much more interesting to me than whether a science story ends up on the main page on a regular basis.

    PR: We are a business magazine and, while I am always fascinated with news and journalism and technology, I have to ask what your business model is? Is it going to be run through an advertising model or perhaps the secure feeds going through the newsrooms? How are you planning to make money?

    SK: It is an advertising model. The actual advertising model is something that is still under wraps – not something we have talked about publicly. I guess all I can really share right now is that advertising is the business model.

    PR: Okay.

    SK: So, this is not going to be like a B2B enterprise where we charge newsrooms licensing fees. The enterprise sales business is not a very interesting business to me. I think there is a much more fundamental web business to be had in the mutually beneficial exchange of data and technology where we enable newsrooms, basically for free, to do interesting things editorially. We can give them access to a larger network of data and the data they create flows into our larger database, and it becomes almost like data cooperative if you will. People put in and take out, and the larger the network, the more powerful it becomes and the more all participants can get out of it. The business model that comes out of that is a very symbiotic relationship with all parties involved.

    PR: In terms of the initial journalists and newsrooms that are signing up, are there many old media big players or is it a fairly decent scattering across the range of old and new media?

    SK: Well, how would you define the old versus new media?

    PR: I suppose it is becoming an increasingly grey area, but I am talking about perhaps the news agencies and newspapers, even though they have digital divisions.

    SK: Well, newspapers have actually been where we have gotten the most traction going in. I see huge potential to work with local newspaper companies, which is what we’ve been doing to a large extent thus far. That is what I was talking about before regarding a new model for distribution of local content. Imagine creating a network out of 1,500 local newspapers in the US. That would be far larger and far more powerful than the sum of its parts.

    Right now each of those newspapers feels vulnerable because it is sitting out there by itself and it no longer has this monopoly distribution model on the web. It’s just another website. And that is the advantage of creating a large network. Digg has the power of millions of users. If Digg was just a hundred people using it, it would not be that interesting. But when you connect all those people from all over the world who use it, that is how it becomes such a force on the web. One of the reasons why bloggers have so much influence on the web is they are the ones linking to everything. And those are the links that Goggle reads and determines what shows up in search results. That’s why bloggers have so much influence.

    Journalists don’t link to anything so they do not influence anything on the web. We want to give newsrooms the power to do it as a network with many other newsrooms, so they can gain more influence on the web. So it is not about supplanting Digg or supplanting Goggle or anything. Right now most media companies put content online. That’s it, and then they put it under their brand, because that’s how people always found it. Most media companies don’t help people do the fundamental thing that they want to do on the web, which is find content. That’s why search dominates the web.

    We want to empower media companies as a collective to do that fundamental thing of helping people find content – that’s where it becomes very counter-intuitive from a traditional media standpoint. Media business was always about sending content out via a monopoly distribution channel. You cannot do that on the web.

    So, you put your local newspaper content on the web where anybody around the world can access it, but anyone around world can access any other newspaper around the world. It used to be that if you wanted news you had your local newspaper and that was it. Now you can get any content anywhere on the web, anytime that you want it. So, how does the brand stay relevant?

    Knox News in Knoxville Tennessee tackled the issue of how they were going to stay relevant on Super Tuesday? They didn’t just publish their own coverage of the Tennessee primary. There were a lot of other people talking about the Tennessee primary, and there was national stuff that they couldn’t cover – it was coming in on the wire so they just linked to it. It provided another huge piece of value for anyone who visited the site for their original reporting. By keeping that flow of headlines going, it is a perfectly complimentary thing, to give readers a reason to come there.

    If you are interested in the election, we are going to give you the most interesting stuff we can find on the web and when you keep coming back we’ll have more interesting stuff. The traditional media mindset says, don’t link someplace else, don’t send people away – you want to keep people on your site. That’s the traditional media principle – keep people locked in here, do not send them away. But I can think of a company on the web that does nothing but send people away. All it does is link to someplace else. It sends people away and sends people away and sends people away. That’s is all it does and, amazingly, people keep coming back again and again and again, and that allows them to make ten billion dollars. When you look at Goggle as the fundamental entity of the web, you realise that if you do a good job sending people away they will keep coming back to you. In many ways we want media companies to be more Google-like.

    PR: Well, that is a fascinating model and I–

    SK: You got me going…

    PR: No, no. That is some great stuff. That’s really what I wanted to attack – how this can play back into the evolution of media and news. There is a lot of navel gazing at the moment and I think it is going to be tools that get people out rather than writing harder or faster or making it more opinionated.

    SK: Yeah. The fundamental problem that journalism has right now is that it’s lost its monopoly distribution. There is no way to guarantee that it gets any attention. So people are putting up the “Digg This” buttons and they are doing search engine optimisation – it’s all good. It is getting more people to that good journalism. But ultimately journalism itself needs to be a force for driving traffic to itself, if you will. And Digg’s not going to do it because Digg doesn’t really care. Digg will drive traffic to any site including newspaper, magazine or whatever site, if it amuses, or to some very narrow definition of what they think is important. There needs to be a force for driving people to important journalism because it is important. You can’t just throw it up there and say, ‘Well, we did it because it has to be done. Why isn’t anybody showing up?’ Because you haven’t bundled it with the real estate section or with the comics. News aggregation is an editorial product. It is content. We’ve set up a basic link-log structure where each bookmark is a link and you can write in a brief article description, which is your thought about why it is interesting, important, newsworthy, whatever. And you string that all together and that is content. That is an editorial product. As much as a piece of original written reporting is. This is a different editorial product. Just like a feature piece is different from a news piece, is different from an opinion column. We have always had different sorts of editorial products.

    PR: That’s very true. Well, thank you very much, Scott. I look forward to continuing the conversation in the near future.

    SK: Great. Thank you.

    Read “The Future of News”, Paul Ryan’s New Media column based on this interview, appearing in the April/May issue of Australian Anthill magazine.