Research chat: Martin Nisenholtz on the New York Times and digital journalism
Martin Nisenholtz is a senior advisor for the New York Times Company and an adjunct associate professor at the Columbia Journalism School. One of the pioneers in the field of digital media, he was senior vice president of Digital Operations for the New York Times from 2005 to 2012. He was the founding leader for nytimes.com in 1995 and chief executive of New York Times Digital from 1999 to 2005. Nisenholtz founded the Online Publishers Association and the Interactive Marketing Group at Ogilvy & Mather.
A spring 2013 fellow at the Shorenstein Center, his research has focused on the history and future of digital journalism. His project, co-produced with fellows Paul Sagan and John Huey, is titled “Riptide.” It is set to be published in the coming months. As part of our ongoing “research chat” series, Journalist’s Resource recently caught up with Nisenholtz for insights on the news business and Web dynamics. Researcher Alex Remington conducted the following interview, which has been edited:
Journalist’s Resource: What key digital skills should all young journalists be acquiring?
Martin Nisenholtz: I co-teach the business of journalism class at Columbia Journalism School now, and probably the answer that most people would give is a kind of range of journalistic skills, including more traditional reporting skills — all the way to the use of computational analysis in journalism.
But I’ll go to a different place, since a lot of folks probably have already talked about that. I think it’s very important that young journalists understand deeply the transitional models, the business model economics, of journalism. In the past, there’s been a very distinct split between the craft of journalism and its practice, and its financing. So you basically had the journalism side and the advertising and circulation side, and never the twain met.
In fact, there’s a very funny story that I can tell, a very brief story, to illustrate this: When I first joined the Times, I met an editor there who had been at the paper for about 30 years. And he was working closely with me on the development of the website. His name was Bernie Gwertzman. He’s at the Council on Foreign Relations now. And I had just come from a meeting with Bill Pollak, who at the time was running the circulation department at the New York Times, and I mentioned Bill Pollak to Bernie. Now, these are two guys who had been in the paper for at least 20 years together, and had ridden up and down on the elevator daily, probably, together. And Bernie had no idea who Bill was. Bill was on the masthead, in fact — but who reads the masthead?
So, it’s just a story about how, even as late as the mid-‘90s, people didn’t even know one another on the business and journalism side. I think that has to change, because the fact is that the business models are undergoing such profound transformation. And so I think it’s important that young journalists really understand the business model implications of digital journalism, and what it means to build, create and sustain a business in this area.
JR: How should their understanding of the business inform the way that they approach their craft?
Martin Nisenholtz: I think it means a couple of things. One is that it guides how they approach the very essence or development of the product. In other words, it’s not just simply the financial side or the commercial side’s decisions, or shouldn’t be, about what gets built and how. The journalists need to be at the table at the creation of the products themselves.
So that’s really what I’m saying more than anything else. I care much less about whether a journalist knows how to run a particular kind of video camera, and much more about whether a journalist can sit at the table and have a discussion about building a new product with people in engineering, with people in finance, with people in advertising, and to kind of think through how this thing is going to be built and financed. Otherwise, it won’t be built and financed. Not in this era. It’s simply not going to happen. So that’s what I mean.
JR: How has the Times’ Web presence evolved over your career? And where do you see it headed in the future?
Martin Nisenholtz: I think that one way to answer that question is to say that the Web presence is no longer just the Web presence. I think it’s fair to say that the Web is and will continue to be the most dominant form, and it may even become more dominant as the app world kind of merges with the Web world. It’s also important to say that, over time, we’ve gotten much better at all the tools of the trade, including video.
Just as a sort of a broad comment, I think the way it has evolved is that at the outset the vast, vast majority of folks in the organization viewed the Web and the digital platform simply as a new distribution channel for the journalism that was being created for the newspaper. You were going to create this newspaper content and copy, and then you’re going to throw it over the wall. At the outset, it was really the 24-hour product. In other words, there wasn’t even any updated news. It was just the content for the paper, thrown over the wall, published on the Web [once a day].
That changed by the late ’90s. We had a very robust, or the beginnings of a very robust, breaking-news operation. And so instead of doing a 24-hour feed, we were doing a continuous feed. We began to do more and more multimedia journalism. And so the journalism side has gone away from being principally stuff that was published in the paper onto the Web. Now I view it as basically the website gets published to the paper once a day. Most of the content that you see on the website by mid-afternoon, or even earlier than that, hasn’t ever been in the paper. So from a purely journalistic perspective, it’s kind of a Web-first mentality now.
I think the second thing that has happened — and it happened, in my view, way too late and in a sense too little — is that the engineering side of the business really came into being. And by that I mean the folks who not only build the publishing systems and help support the publishing of the content on the Web, but the folks who are developing applications around the content, and making the journalism much more, in a sense, valuable to people in a digital format. And that comes in many flavors. I think we now at the Times have a significant engineering capability. Is it like a Silicon Valley tech company? Of course not. People suggest that journalists need to become engineers. I think that’s an overstatement. But I think we at least have the foundation now to create new products in a way that is much more application-focused than they were in the past.
JR: What are the implications of that for the Times’ digital presence going forward into the future?
Martin Nisenholtz: I think the key implication is that this stuff gets much more useful. I think that there’s a distinction between a one-way broadcast mentality with respect to content, and a services mentality. And I don’t think they’re at odds with one another. I think it’s fine for people to come onto the website and other digital products — at the Times, the Wall Street Journal, the Washington Post, wherever they are — and read them. There’s nothing wrong with that, that’s what they’re there for, in some ways.
But they also need to be used. So the content needs to be searchable, just as a basic premise. You need to be able to go in. The Times, as an example, has one of the deepest travel archives on the Web. We’ve got wonderful content, if you want to go to a particular place. That’s not just a matter of publishing the content, it’s a matter of finding it, and having it available in a place that’s useful to you. Same with our movie reviews. The Times has reviewed every meaningful movie that’s been made since the inception of film. You should be able to go back and find those reviews.
So, search is just one example of a Web service that people use in order to make the content more valuable to them. And there are dozens of those services. So you need engineers, and engineering talent, to be able to bring them to life.
JR: What should Web producers and editors keep in mind regarding online search? What should they be thinking about in terms of social media?
Martin Nisenholtz: There are a whole range of techniques. The beauty of the New York Times’ content is that because we had relationships that predate the Web — that go back well before the invention of the Web — our content was very well-structured, tagged and bucketed for search. Essentially, the Times was one of the inventors of a searchable news archive. The Times Information Bank was the precursor to Nexis. In fact, it became Nexis. So when I joined the company in ’95, the content was already extremely well-structured.
This was true on an article basis. What I mean by that is that you could go find an article. Now, everyone knows that something like “36 Hours” — which is a feature that we do, again, back in the travel area — is an article. But it may have things for people to do listed in the article, it may have places to stay in the article, it may have restaurants — that’s much harder [to surface]: Parsing that article and breaking it down into its component parts, and saying, “OK, if I want to go to London I don’t necessarily want to read the whole ’36 Hours,’ I just want to see where the ‘places to stay’ part is.” That all becomes a more difficult part of the problem. Being able to actually read that specific content and structure it to do so.
We’ve gotten, I think, quite good over the last ten years at structuring our content in a way that allows, in the case of Web search, Google’s spiders to see the content and index it in a way that makes it available to people, and that’s part of the job of those folks in the context of search. With respect to social, that’s a whole different can of worms. Googlebot looks at the content one way. But human beings — who we would like to have tweet the content, or share it on Facebook, or do whatever they do with it in social settings — that’s a different thing.
In fact, one of the more interesting points that I think Jonah Peretti of Buzzfeed made in his interview with us during the “Riptide” research project, was that in the past, you could create a page — a very, very simple page — off of an existing article and get more traffic to that page through Google just because you were optimizing it. Per my point about the editors and search, in social, the shared article is really the original article. People aren’t going to necessarily share something that is a sorry shadow of its original self. Right? They’re going to want to share the original piece. And so, Jonah’s point was, that in the social era, there’s a kind of resurgence, or there could be a kind of resurgence, of original journalism, because in his words, “Googlebot is dumb.” Now, it’s not dumb, but it’s crude in comparison to the judgment that a human being might use over a particular article. So I think that’s one of the benefits of the social web. I think Jonah’s right about that.
We’re at a stage now where the more traditional journalistic operations still have so-called “social media editors” that kind of massage and use the social networks pretty much just for the enhanced distribution of content. I think the difference between what the traditional folks have done and what Buzzfeed has done, obviously, is that Buzzfeed was built from the ground up to be a social news company. In other words, a lot of the content is written specifically to be shared. And that has its pluses and minuses. I’ll let the audience judge them based on their use of Buzzfeed. But the point is, that business, that journalism entity, was built for that purpose. So I think you can see the contrast between how editors and journalists at traditional places might view social media versus something that’s built from the ground up for it.
JR: What are the most important numbers to you in terms of Web analytics? What are some underappreciated aspects of analytics?
Martin Nisenholtz: I think numbers are really important, in large part, as a function of your business model. We know, for example, that there has been a somewhat increasing trend toward the atomization of content. So, Twitter, as an example, becomes your newsreader. And that’s fine, at some level. But if you’re in the business, as we are — as the Times is — of building a subscription base and a loyal audience, you want to continue to measure the strength of your home page.
And you want to measure the strength of your home page for two reasons.
One, returning visits to the home page show a kind of loyalty to the brand. That is represented by more than just the view of an article off of Twitter. And it translates in theory, or should, to the possibility of a real paying subscriber. Because, as we know, paying subscribers use the sites that they pay for much more than non-paying subscribers. That’s just logical. So that becomes a very important metric.
In addition, the home page itself is the point at which you have the most traffic and the most attention, in a sense, on one page. So it becomes a significant advertising business. In most websites, from Yahoo on down, the home page is the single largest point of revenue production in the advertising area. Measuring and understanding your ability to maintain a kind of coherency around a point of entry is, I think, not quaint. I think it’s an important thing to do. Now, others would disagree. A business model like the London Daily Mail — I’m not part of this team obviously — has kind of the opposite approach, in which you’ve got tons and tons of traffic and users and you’re really, really super-optimizing your product for search and for social media and all that. So, the side doors become really, really important. It’s not a subscription model; it’s based on advertising revenue to pages. It doesn’t matter so much that your users are coming to you through the home page — though I think even they want significant home page traffic, because, again, that point of entry becomes a very significant place for advertisers to be. But all I’m saying is that in that model it may be a bit less important. And it may be just more important to have pure mass. The ability to go to the advertising community and say, “Well, we have 60 million, or 70 million, or 80 million uniques, and we have these many page views that we generate.”
I still think that the “90/10 rule” really does apply in most news sites. So, by that I mean 90% of the usage is driven by 10% of the users. And I would bet that’s true, around those numbers, for virtually every news website out there. The new subscription models change that somewhat. But it’s still somewhat true.
JR: How do you balance the need to grab attention with the need to maintain serious professional standards?
Martin Nisenholtz: I’ve never viewed that as a particularly difficult thing to manage.
That’s been a controversial thing from the beginning. I think there was the sense, maybe 10 years ago, 15 years ago — when the whole drive was for eyeballs — that Web news sites were going to go the way of television. In other words, it would become a ratings-driven business, and editors would have to sit in front of screens and basically optimize for traffic. And that’s true for some entrepreneurial news operations; I’m not going to mention any names. People are sitting there looking at Chartbeat, looking at which articles are trending, and optimizing those articles for usage. They’re just basically building a kind of editorial overlay based on what is popular and what is trending.
I don’t think that you have to do that. And I think particularly now that a lot of the more robust journalism entities — and by that I mean the places like the Wall Street Journal, the Financial Times, the New York Times, soon the Washington Post — places with big newsrooms, in essence, are now charging for content in some form or another, it really changes the analytics and the measures. It focuses you much more on loyalty and attractiveness than on purely a “ratings-driven business.”
On the other hand, you are in an extremely competitive world. I think great editors have always been capable of, and really great at, creating a kind of wonderful, serendipitous blend of things for people to see. If you talk to users of websites even now, this notion of highly targeted personalization isn’t really that attractive, because people really like the idea of serendipity and seeing things that they wouldn’t necessarily see on their own. The idea of quality is a question not so much of serious journalism and the Baghdad bureau as it is a mix of surprising, interesting, wonderful things that at some level you’re willing to pay for. I don’t think, in this world, that you can force people to eat their broccoli. I think you have to bring something out that’s really compelling. That doesn’t mean that it can’t be journalistically robust and serious, but it has to be interesting at the same time.
JR: Speaking of personalization, what do you think about “filter bubbles” and whether or not we’re becoming siloed in hypertargeted bubbles or bubbles of our own choosing? The idea is that everyone who is, for example, in the engineering community, or everyone who identifies with a certain political party, will end up in the same digital bubble. Do you see that as a serious issue?
Martin Nisenholtz: I do see that as an issue [but] I don’t know what the trends are with respect to how people are actually using these services, for example, by demographic cohorts. So, for example, are baby boomers still going to destinations and seeing the serendipity, versus people in their 20s who might be more apt to use, to use your term, a filter bubble? I just don’t have any information about that.
JR: And what about media polarization — or perceptions of bias? Several years ago, New York Times public editor Dan Okrent wrote his column, “Is the New York Times a Liberal Newspaper? Of Course It Is.” Certainly, many readers perceive the Times as liberal, which presumably affects the way that many readers will approach the news content. Does that affect the way that you attempt to target new readership, depending on where they fall?
Martin Nisenholtz: I don’t really want to comment on the New York Times’ opinion or editorial posture. I’ll let Arthur Sulzberger do that. That’s his job. I could just say, speaking personally, I don’t believe that the news content of the Times — and this is just my personal view — is either liberal or not liberal. I think that the folks in the newsroom do the best possible job they can to play it as straight as they can, and I believe, by the way, that’s true at the Wall Street Journal as well. I don’t think people sit around in the newsroom at the Wall Street Journal and plot about how to provide a conservative slant to the news.
I understand Okrent’s point, which is that many of the people who work for the New York Times, and, by the way, for the Wall Street Journal, live in Manhattan, tend to have a more urban perspective on life, and may, in that context, be a bit more liberal than not. I don’t know whether that’s true or not. It may be. But I’m not sure that from a purely journalistic perspective or a news perspective, it needs to matter all that much. I think if you look, for example, at the Times’s coverage, its recent coverage of the IRS scandal, you’d be hard-pressed to say, “Well, jeez, the Times is sitting around trying to figure out how to support Barack Obama.” We’ve been as tough on that as you can possibly be, in my opinion.
JR: What will the New York Times look like in 2020?
Martin Nisenholtz: That’s soon. One of the things I’ve learned is the idea of so-called Amara’s Law: The idea that things seem to change more slowly in the short term but much more profoundly in the long run. This was one of the themes of our “Riptide” research at the Shorenstein Center, as well. It’s true.
So when you go back just six years ago, in 2007, before the financial crash, there were a lot of things that were different back then. Facebook was not really a factor so much at that point; Twitter wasn’t really a factor. But the fact is that, looked at over the long term, consumer behavior around news hasn’t changed all that much.
Obviously, it’s changed with social media in the sense that people are getting to the content in very different ways. I get that. But we had roughly a million subscribers to the paper in 2008, the daily paper, and we have the same number now. Roughly 1.8 million, or 1.7 million, in total.
So, in 2020, I’d say that it’s likely to be different, but not all that different.
Keywords: training, research chat