On-demand Media

Online newspapers: deep links and segmentation

I've been trying to put my thoughts in order after my previous two posts. To recap, I was musing on the relative value to news sites of users who arrive straight to an article via links found on other sites, versus users who start by visiting the site's front page, and whether it is ever worth sacrificing a bit of the first in order to get a bit more of the second. Here's an attempt to put some of the issues in standard marketing-speak.

Your primary product is your articles. You make money when people read them. And for this to happen you need distribution. Distribution here means putting a link to your article in front of prospective readers so that they can click on it, if they want (it does not mean physically transmitting the content over the internet.) You have two main ways of distributing your content:

• The first is to go direct to consumer. This means persuading users to come to your site – to your front page – on a regular basis to check what you have to offer. This the brand-driven repeat-purchase idea that marketers have always understood. The ingredients are an audience segment, a targeted product that resonates with it, and a brand that anchors the relationships of trust that must be built before 'repeat purchases' can be counted on.
• Alternatively, you can go via channels that put your wares – links to your articles – where their users are. The obvious approach here is to pay your channels – e.g. portals – to do this, but that is only a small part of the story. Whenever a blogger, a journalist at another site or a Myspace user puts a link to one of your articles sh/e is acting as a channel for you. For online news, channel management is not only a business-to-business thing.

Everything so far is common knowledge. What follows may not be (and is somewhat conjectural – corroborating or contradictory evidence would be appreciated):

1. Most sites get many more users (although not necessarily more hits) via indirect channels than via their front page.
2. You can't aim to control your channels. There are too many of them. And they are your readers, so you should be nice to them.
3. Your direct channel (your front page) plays a crucial role in driving indirect reach. If Google News links to one of your articles that is because it monitors your front page. If a blogger publishes a link to one of your articles that is because she found it on your front page, or on another site whose author saw your front page (or on a third site whose author found it on a fourth site whose author saw your page, or on a fifth site... you get the point).
4. The more viral distribution you get (the mechanism in the previous point), the more your articles will appear in Goggle results and the more reach you will have.
5. Most of the users you get via indirect channels will never establish a direct relationship with your brand and become regular visitors to your front page. They are from a different segment. So they won't cannibalise your direct segment. They bring incremental, not substitutive reach and thus are nothing to worry about. You should welcome them.

This has a number of implications:

• Even if most of your distribution is viral, all of your users come via your front page in an indirect way. Neglect your front page and you neglect all your readers. Unless you are a newswire, you can't do news without being a portal.
• How your core readers – those who use your front page – perceive you influences how everyone else will. How respected they are influences how respected you are (no least by Google). So choose them carefully and earn their respect.
• The more indirect a viral link is (in the sense of point 3 above) the more time has passed since you first published your article, and the likelier it is that it is being linked-to for historical purposes and not as news. So make sure that your articles stay live permanently at the same URL even after you've stopped linking to them.

Finally, how does this relate to the NYTCo / Gatehouse dispute I wrote about in a previous post? Gatehouse's demand that the NYTCo stop deep-linking to its site suggests that my point 5 above did not apply in that case – i.e. that the users the NYTCo was sending to Gatehouse's Widked were from the same segment that Wicked's front page was trying to target. Still, why did Gatehouse object, if all of these users were as likely to distribute links to Wicked? My conjecture: because Gatehouse (i) thought it would make more money from these users if they came to its front page, which it thought they might do; and (ii) it wanted to keep a direct relation with them so it could better influence the viral process by which these users would bring others who would never visit the front page. A cash-positive decision in the long term? Time will tell.

 

TV to follow newspapers?

Jeff Jarvis reflects that newspapers' current woes will soon come to broadcasters. His thinking is that both newspapers and TV stations are guilty of the same sin: taking advantage of an oligopoly of distribution that is now losing relevance.

Perhaps. But I wonder if there is also another force at work. As I mused at the end of my previous post, newspapers and broadcasters also share a curse: having to 'carry' or 'reproduce' all the content they offer, which means they have to own it or licence it. Of course they also get to monetise it, but sometimes that struggles to cover the costs.

The magic thing about the web is that, thanks to the link, you get to offer stuff to your users without having to carry it – and hence without having to own or licence it. Of course you don't get the upside, but you can get some money just for listing stuff, and you make your users happy and they come back. So you can cut back on content costs and only do the content that you do best (and profitably), and link to the rest (Jarvis).

It's not just that newspapers and broadcasters can't play this game. Their online properties can't do it either, because they already own tons of content (because their print/broadcast siblings need it). So they can't compete on the same terms as the new online-only guys. And if this is the new game (is it?), the old guys could be toast.

How strong is this force? I'm trying to find out. Any evidence either way would be very appreciated.

The Gatehouse / NYTCo settlement

The recent settlement between Gatehouse and the New York Times Co over deep linking and aggregation has been noted far and wide. Most people seem to think that Gatehouse's complaint was unjustified and that by settling the NYTCo set a bad precedent. Some also think that Gatehouse was acting against its own interests because more links mean more readers and more money. I'm not so sure.

To summarise (see a good Q&A here), back in December Gatehouse sued the NYTCo over the way in which NYTCo-owned boston.com linked to articles in its Gatehouse-owned rival wickedlocal.com. Among other things (full claim here), Gatehouse complained that

1. boston.com was violating Gatehouse's copyright by reproducing its articles' ledes verbatim
2. boston.com was hurting wickedlocal.com's revenues by deep linking to its articles (i.e. allowing readers to bypass wickedlocal.com's front page and its ads and go straight to the articles)
3. by listing wickedlocal.com's headlines (with attribution) next to other content it was diluting wickedlocal.com's brand and
4. boston.com was giving its users the false impression that there was a syndication deal between Boston and wickedlocal.com

This involves so many issues that I won't even attempt to cover them here. But as Dan Kennedy noted at the time, "the future of online journalism" could hang on this case. Aggregation is at the heart of many of the new business models for news that many people are exploring, and whatever rules emerge in cases like this will dictate what business models are and are not possible. An angry Jeff Jarvis shouted: "if you can't stand the links, Gatehouse, get off the web."

Under Monday's settlement, NYTCo agreed to remove all the headlines it copied from wickedlocal.com, and to stop automatically aggregating content from any site owned by Gatehouse. Although this was settled out of court, it left people complaining that the NYT caved in to an "absurd" suit that Gatehouse had little chance of winning. Techdirt lamented the "cluelessness that goes into actively turning away the kind of traffic a major media publication can provide", and Mathew Ingram worried that "every settlement like this one weakens the defences around the entire structure of the Web, in which linking and quoting — in some limited, representative way — is a fundamental principle".

But the fine print suggests that this was not a one-sided victory. Each party paid its own legal fees. And despite the prohibitions above, the agreement states that "nothing shall prevent either party from linking or deep-linking to the other party's websites". In an interview given after the announcement, Gatehouse's COO Kirk Davis clarified his position somewhat. The following paragraphs (by Zachary Seward of Harvard's Neiman Lab) are revealing:

The Globe's Your Town sites could link to specific articles or blog posts published by GateHouse as long it doesn't quote the headline or lede, according to Davis. "We don't have an issue with deep linking," he said. I asked Davis if it would be OK for a Globe blogger to quote a few grafs from an article in one of GateHouse's community newspapers. "I would agree that's fine," he said, adding with a huff, "That takes a little bit of human effort."

There's the key distinction: automation vs. human effort. The finer points of this settlement will be ironed out in practice, but this definitely strikes a blow to automated aggregation.

I think this may be a crucial point. But to explain this I first need to clarify the terrain. Two things:

First, online newspapers get their advertising money at two moments.

• One is when someone reads their content – that is, their articles
• The other is when someone sees their front page (if they advertise on it). Front pages are much more productive (have a higher 'yield') than articles because they get a lot more traffic than any individual article.

Second, online news sites have two types of users.

• 'Nice' users know the site by name (its brand), visit regularly, and always start by the front page. Once they arrive on average they go on to read more articles than the bad, 'nomad' users below
• 'Nomad' users arrive via a 'deep link' on another site like Google News or a blog that takes them directly to an article. They read what they came for and then go away leaving behind far less money. Rude.

That may be true, you may say, but 'nomad' users are still worth something. They read articles and generate revenue. Google News is giving Gatehouse that money for free, and they complain! Why?

Well, boston.com is not Google News. It is a direct competitor of wickedlocal.com, operating in a hyper-local market with only two players. Every 'nomad' user that wickedlocal.com gets from boston.com is a user who just seconds before visited boston.com's front page and was quite 'nice' to them. If you are Gatehouse, you would be wondering why they couldn't be nice to you instead. And as you ponder that, you may conclude that they are nicer to boston.com because boston.com treats them better than you. And that you are helping boston.com be nice to its users by letting them link to your articles, in addition to whatever else boston.com has to offer. They have a superior product and it's all thanks to you (or at least partly). So what do you do? You stop being nice to boston.com and tell them to stop linking to you. The key question is whether the lost revenue from fewer 'nomad' users from boston.com will be made up by more 'nice' users. That's a difficult call to make, but it may be that Gatehouse concluded that it would.

I agree with the sentiment that Gatehouse's reaction feels somewhat against the grain of standard practice in the web. But still, a concession must be made to its claim that there is a difference between including a link after an editor has judiciously decided that an article is worth mentioning (what Joshua Benton called "curation", to which Gatehouse doesn't object), and the wholesale copying of wickedlocal.com's running orders by a direct competitor (the claim made explicit reference to using RSS feeds). In essence, Gatehouse is saying that you can't just go and replicate the way it aggregates its content because when you do that you are taking its most precious asset.

(More abstractly, the claim is that an aggregation of links is itself valuable content that should be subject to copyright. But then, could you own the copyright to an aggregation of links to content that isn't yours? Could you, e.g., lift the aggregation in Google News and display it in your site? I think it is a matter of time before this sort of thing is tested in the courts.)

Having said all this, I'm not sure I would have reacted as Gatehouse did. A simpler response would have been to reciprocate in kind by systematically linking to all of boston.com's best content from wickedlocal.com's front page. Your front page would keep its pulling power and your users would be better off. Sure, boston.com would get to keep the revenue from further page views, but you can't expect not to lose some revenue after the arrival of a powerful new competitor.

Finally – and here I really am speculating – it may have been that Gatehouse couldn't do this because it had too much content. If it has its own version of every story that boston.com covers, then it couldn't link to wickedlocal.com's articles without undermining its own content to some extent.

This scenario is being faced by all online newspapers that are web editions of print papers (not wickedlocal.com's case): you can't follow Jeff Jarvis' dictum to "cover what you do best and link to the rest" because you already cover everything, because you need it for your print edition. I've been wondering for some time if this is the key structural weakness that all print newspapers have when competing with online-only rivals in the web. The problem may affect all players with an old-media presence, not just newspapers and not just text publishers. I plan to write a post on this in the near future.