"The problem is how do we get back to that glorious bundle that the paper did so well?" Bezos asked at a question-and-answer session with Post journalists.
Bezos lauded the "daily ritual" of reading the morning newspaper over coffee. "That daily ritual is incredibly valuable, and I think on the Web so far, it's gotten blown up."
But that daily ritual got blown up for good reason. Trying to recreate the "bundle" experience in Web or tablet form means working against the grain of how readers, especially younger readers, consume the news today. In the long run, it's a recipe for an aging readership and slow growth.
Sun Microsystems co-founder Bill Joy once said that "no matter who you are, most of the smartest people work for someone else." That's why the smartest executives, especially in the technology sector, are constantly looking outside the boundaries of their own firm for good ideas.
Joy's Law has a corollary for the news business: No matter how good your news organization is, most of the best journalism is being done somewhere else. That's because no publication, even storied outlets such as the New York Times or The Washington Post, can hope to hire a majority of the world's most talented journalists.
And this is why the smartest readers have increasingly eschewed "bundled" news outlets in favor of third-party aggregators that provide them with links to the best news from around the Web. In the early years, the most tech-savvy users used RSS readers. Then news aggregators like Google News, Digg and Reddit began to appear.
In the past five years, aggregation has been democratized by social media. A growing number of younger readers don't actively seek out news at all. Instead, they read the news that's recommended to them by friends on Facebook and Twitter.
That's more convenient, because most young people are spending time on Facebook and Twitter anyway. More importantly, it serves as a finely honed news filter. Probably the best predictor of what news stories you're going to want to read is what news stories your friends and colleagues found interesting.
If this trend continues, and I think it will, then the future of news is one of radical unbundling. A large share of every publication's traffic will come from referrals from third-party aggregators, and the key to success will be crafting content that performs well on these sites. Readers who prefer to read a publication's news "bundle" from front to back will represent an aging and shrinking demographic.
The power of Facebook
In yesterday's Q&A, Bezos was dismissive of this approach to news. "If our readers read a couple of articles through the Web or Google News, a couple per month, that's a small business," he said. But this ignores another corollary to Joy's Law: No matter how popular your news site's home page (or tablet app) is, most readers are going to rely on someone else's site to decide which news stories to read. There are about 200 million adults in the United States. If every American read two Washington Post articles per month, that would amount to 400 million monthly page views. I'm not privy to The Post's own traffic statistics, but for comparison, the most popular blog on the Los Angeles Times Web site, LA Now, got less than 20 million page views in February, its best month ever to that point. Two visits a month from each American would represent a large boost in the L.A. Times' readership.
And there's no reason to think that's an upper bound on the amount of traffic that can be generated by links from other sites. The average American reads a lot more than two articles per month, so a publication with a knack for producing articles that people like to share should easily be able to entice the average reader a lot more than twice a month.
This is more than a theoretical possibility. Indeed, Bezos himself hinted at the importance of social media traffic in his comments at The Post this week. At two different meetings, Bezos cited Max Fisher's masterpiece "9 questions about Syria you were too embarrassed to ask." I don't know exactly how much traffic that post generated, but we can get a ballpark estimate from the fact that the post has received 578,000 Facebook "likes."
In other words, the post was wildly successful because it appealed to hundreds of thousands, perhaps millions, of readers who wouldn't ordinarily visit Fisher's foreign policy blog or even The Washington Post's home page. The problem with the "bundle" strategy, then, is that it encourages writers to think small. It asks them to cater to the hundreds of thousands of people who read The Post every day, ignoring the vastly larger universe of readers who might be enticed to visit the site from elsewhere on the Web.
Headlines are advertising
And that's connected to the other major challenge Bezos identified. "Every story gets rewritten in 100 places," Bezos said. "Should we stop doing investigative journalism because it's unrewarding and other people copy it? No, we have to figure out how to get back to that bundle."
But those 100 rewrites don't all get the same amount of traffic. Typically one or two of them will catch the imagination of Reddit and Twitter and get more traffic than all the other rewrites put together. And the original version of a story has some inherent advantages over the rewrites. It's up first, of course, and its authors also have a lot more lead time to hone it into a polished, eye-catching package. In principle, then, there's no reason in-depth investigative journalism shouldn't be able to generate significant traffic.
The problem is that traditional news organizations, while good at gathering the news, are often bad at selling it online. Take, for example, this New York Times scoop from last year about Target using data mining to predict what products customers want to buy. The Times gave the piece the boring headline "How Companies Learn Your Secrets." Kashmir Hill at Forbes did a rewrite of the story with the much catchier headline "How Target Figured Out A Teen Girl Was Pregnant Before Her Father Did." The result? A lot of the social media traffic went to Forbes, netting them 2.1 million page views.
There's no reason the Times couldn't have titled its own article "How Target Figured Out A Teen Girl Was Pregnant Before Her Father Did." And if it had done so, it probably would have gotten most of the traffic that went to Forbes. But the Times is steeped in newspaper culture, with its tendency toward understated headlines. So great reporting didn't get the financial reward it should have.
At root, the tendency toward boring headlines flows from thinking of a newspaper as a bundle. Customers buy entire newspapers, not individual articles. So by the time the reader has opened a newspaper, he's already a captive audience. That gives newspaper editors little reason to write flashy, eye-catching headlines.
In contrast, people read online news one article at a time. Every article is competing with thousands of other articles for the reader's attention. And that makes eye-catching headlines much more important. "How Companies Learn Your Secrets" may have been a good summary of what the Times story was about. But "How Target Figured Out A Teen Girl Was Pregnant Before Her Father Did" was a lot more likely to get people to click on the story and share it with their friends.
To be sure, The Post has hundreds of thousands of loyal readers who do like reading the news as a bundle, and The Post needs to continue serving those readers well. But those readers are largely the same readers who prefer reading a paper newspaper. It's an aging demographic, and focusing primarily on their needs is a recipe for decline.
A growing number of younger readers want to read their news one article at a time, just as they listen to music one song at a time. So if newspapers want to thrive in the long run, they need to embrace the unbundled future and treat every article as a product in its own right.
Correction: This article originally stated that the Los Angeles Times website got 20 million pageviews in February. In fact, the figure was for LA Now, the LA Times's most popular blog. We regret the error.