(See the first part of this post)
What the Journal does have that nobody else does is more than a million paying subscribers to its Web site, giving it what I’ve very roughly estimated at $60 million to $70 million a year in subscription revenue.
But—and this is critical—that revenue is not all the Journal gets from its subscribers. It gets ads from them, too, and I’m not privy to the numbers, but I’ll bet you that each spends far more time on the site and sees far more pages than the average non-subscriber. In other words, the subscribers account for a disproportionate amount of the advertising revenue, too.
The point is, you can optimize your search engines all you want, but if you have a high fixed-labor-cost structure like a newsroom, that ain’t gonna cut it, and your core readers are who you need to focus on, not adding marginal ones that add near-zero revenue.
There’s been a little mini-trend of sanity amongst online publishers talking about focusing on their core loyal visitors rather than chasing after the sugar high of a surge of junk Digg traffic or whatever.
Michelle McLellan reported that Slate’s editor David Plotz, in a Mizzou speech recently, said that Slate has seven million unique visitors but is shifting its priorities to focus on just 7 percent of them. Why? Because they’re the loyal readers who love the site’s stuff (or hate it so much they can’t resist it—shout out to Audit readers Mike H and Mark Richard!) and are by far the most valuable to it.
“Until now we’ve been selling to the mass audience. Now once you have this ability to target you can really target your core audience… This creates strong incentive to create durable journalism,” Plotz said. “That one curious reader is worth 50 times the value of the drive-by reader. The person who makes a commitment to your brand, if you’re a quality brand….. if you can get those readers, a smaller set of readers, who come to you three or five or 10 times a week, you don’t have to go after that huge other set of readers.”
This is a critical point. You can’t be everything to everybody. Don’t try to be. MinnPost’s Joel Kramer is making similar noises.
But on most websites, ours included, the vast majority of unique visitors are passersby. They come through a search or a link from a blog, and they visit the site precisely once, usually for a quick glance at one page. In many cases, if you ask them 30 seconds later which site they just linked to, they won’t remember. It would be folly for us to build a business plan around consumers like that.
I’ve been asking for numbers like these for a long while, so thanks to Kramer for providing them:
According to Quantcast, 168,000 different people visited MinnPost during the past month (as measured on Oct. 28). This is a somewhat smaller number than unique visitors, because Quantcast aims to count a person only once even if he or she visits from, say, both office and home.
About 71% of these visitors to MinnPost in the past month were passersby, Quantcast reports, and they accounted for 32% of the visits to the site. Another 29% of the visitors are called “regulars,” which means they visited the site at least twice but fewer than 30 times, and they accounted for 51% of the visits. Finally, there are the addicts — fewer than 1% of MinnPost’s visitors — who visited at least 30 times each and together accounted for 17% of the visits.

This analysis is effective proof that search engines don't pay off purely as a distribution channel. But they're also a marketing channel. What's the monthly benefit to the WSJ's brand and brand-awareness of showing up in Google search results?
#1 Posted by Michael Andersen, CJR on Tue 10 Nov 2009 at 05:09 PM
"What about the top 3 percent or top 5 percent of visitors? If the top 1 percent of loyal visitors account for 17 percent of page views, as they do at MinnPost, here’s guessing the top five percent make up some half of page views."
The "two audiences" problem is very real, but this assumption is far afield. Newspaper site audiences aren't shaped that way.
I would caution against newspapers making ANY projections based on MinnPost.com's experience. It's not a newspaper, it does not have significant market penetration in its own market, and it's still only two years old.
#2 Posted by yelvington, CJR on Tue 10 Nov 2009 at 08:55 PM
Good, thoughtful analysis. But I see two holes in it. First, right now, as I understand it, advertisers are far more interested in reach than engagement. This might (and should) change as they come to terms with the point you're making. But for now it's the reality.
Second, that fly-by traffic isn't totally worthless. Some of the readers who come from Google can be converted to regulars. I agree that loyal readers should be our focus, but we need to grow that audience - and one cheap, effective way to do so (compared to the old ways - direct mail, advertising) is to hook them thru search and land them once their on the site with valuable content, interactivity and easy registration. So we should be focused on uniques not as a bottom line, but as a starting point; then we should measure how effective we are at bringing them back as loyal readers.
#3 Posted by Brandon Copple, CJR on Wed 11 Nov 2009 at 01:15 PM
Michael, Brandon,
I'm not saying the fly-by traffic is totally worthless—I'm trying to put it in perspective. That stuff certainly has some direct monetary value gained from increasing the traffic numbers, however minor a role they do play. I'd argue, though, that that's in part offset by, at least in the case of a metropolitan or regional publication, the dilution of the value of what an advertiser is paying for.
Certainly there's hard-to-quantify non-monetary value to be gained from Google traffic, particularly from marketing as you say. I doubt it's a whole lot though, at least for established publications. I'd think you're most likely to become a rabid everyday reader of something you know about via who you are and who you talk to. If you live in Tulsa, you're going to know about the Tulsa World or Urban Tulsa just by living there. If you don't live in Tulsa (or have a connection to it like I do), you're not going to start reading either just because you happened on to that story about the penis-pump judge that one time.
On your first point, Brandon, I don't think you can have one without the other. If you're coming via Digg and staying an average of 10 seconds on a site, no advertiser cares about you. I'd also reckon it's much easier to sell your core audience to advertisers. You know who they are, unlike transient traffic.
Steve, I would hope newspapers wouldn't make projections based on MinnPost's experience. They have access to their own data. I don't. If you or anybody else would like to show me some, you know where to reach me.
#4 Posted by Ryan Chittum, CJR on Wed 11 Nov 2009 at 06:01 PM
I'm not sure if I would pay money for an online newspaper that still showed me ads. I've grown accustomed to spending money (at salon.com, flickr, and more) so that I don't see advertising.
Of course, I'm probably in the extreme minority
#5 Posted by Anca, CJR on Wed 11 Nov 2009 at 06:48 PM
Anca,
This loyal/junk thing doesn't necessarily mean pay models are the right answer. It can be used just as easily for a non-pay strategy. I just think it's an argument in subscriptions' defense.
#6 Posted by Ryan Chittum, CJR on Wed 11 Nov 2009 at 07:00 PM
There are different levels of demand for entertainment and information.
Some just want to watch. These are the "fly by" folks you are reference. They want pre-digested, relatively superficial information.
Some believe they have something to contribute. These people will pay to get all the information they can, to contribute, to meet others who also care this much, and to critique.
Mass media has met the needs primarily of the ones who want to watch.
There are few examples of media designed to meet the demands of the more active participant. So drawing any conclusions about the potential market for those who will pay is not very helpful.
Dan Conover is on to something when he says that more thoughtful structures for designing information is needed: http://bit.ly/4xWK8J
Katherine Warman Kern
@comradity
#7 Posted by Katherine Warman Kern, CJR on Thu 19 Nov 2009 at 02:23 PM
We are probably a pretty average small newspaper site http://www.nashuatelegraph.com and the top 2% of our readership generates almost half of our page views. And, 40% of those top readers find our site via Google.
So - don't underestimate the extent to which people think Google is part of their Web browser and use the search box as a URL short cut. The top 3 - 4 incoming searches for us are all variations on "nashua telegraph." Go figure.
Damon Kiesow
#8 Posted by Damon Kiesow, CJR on Tue 24 Nov 2009 at 08:34 AM
Damon, thank you so much for that data. That 2% of readers generate 50% of pageviews is incredible. And the search term thing is interesting, as well. I suspect that even if a newspaper yanked it's stuff from google (something I thunk is very unlikely, to be clear), it could have a front-facing home page, even if a redirect URL, that would be No. 1 in search results
#9 Posted by Ryan Chittum, CJR on Tue 24 Nov 2009 at 12:22 PM
Ryan -
I blogged this with a chart illustrating an average month's traffic:
http://kiesow.net/2009/12/04/where-does-the-paywall-go/
Damon
#10 Posted by Damon Kiesow, CJR on Sat 5 Dec 2009 at 08:34 AM
Ryan -
I blogged this with a chart illustrating an average month's traffic:
http://kiesow.net/2009/12/04/where-does-the-paywall-go/
Damon
#11 Posted by Damon Kiesow, CJR on Sat 5 Dec 2009 at 08:35 AM