Tuesday, November 24, 2009

How Demand Media's Business Model Can be Applied to Niche Sites

Demand Media has advertising-driven content down to a science. Instead of creating content for the Web and hoping that it generates revenue, the company works backwards by determining how much revenue each piece will generate before anything is produced.

The company uses a series of algorithms to pick through keywords that people are searching for on the Web and aims to create content unique enough to rank highly in those search results. It also determines how much advertisers would pay to be next to that content.

This is much different than simply using analytics to shift stories around on a home page or testing which headline will draw more readers. Demand is all about the dollars.

News organizations looking to create profitable content on the Web can see that Demand Media's model does make money -- although it forgoes editorial judgment and a journalism process. Yet news organizations could apply lessons from Demand's approach to their own companies, not for standard news operations, but for niche sites that are focused on reader demand and generating revenue.

Demand Media is focused on "service journalism," said Adam Weinroth, the company's vice president of strategic marketing. "This is the kind of content that is evergreen, and includes formats like guides, how-to's and tips."

Besides the company's method of choosing stories, the other part of Demand's strategy is in how it gets its content. Rather than try sell ads to support content that costs a particular amount, the company has dropped the cost of production to make sure it can be supported by what advertisers are willing to pay.

For a company that relies on more than 10,000 freelancers that crank out some 4,000 videos and articles a day on the cheap, the formula works. Demand is expected to bring in $200 million in revenue this year, paying out $17 million to its Demand Studios content producers. Of course when content producers are being paid $15 an article and $20 per video and scrambling to create multiple pieces a day, the quality of the content will likely suffer.

Demand's content goes up on one of its brand or partner sites, like eHow.com or The Atlanta Journal-Constitution's site. It uses sites like YouTube, for which Demand Media is top producer, to share and spread its content, getting 2 million impressions a day with video titles such as "How to change golf spikes" or "How to build a robot our of a box." And its YouTube channels are popular. Expert Village, the most popular, has 448,000 subscribers, with some 14 million channel views.

Ken Doctor, a news industry analyst and author, said Demand's approach is similar to what newspapers and other publishers have done before, creating ad-friendly content such as bridal sections that they know advertisers will pay well to be in. Demand Media has taken that up a few notches. "I give them a lot of credit for developing a very good technology," Doctor said.

News organizations could employ a strategy like Demand's on their subject-oriented niche sites, using the revenue to support other news efforts (just like they do with bridal sections). The niche sites would have to focus on topics such as travel, health and personal finance, which support high-dollar ads, Doctor said.

News organizations already know what content on niche sites such as Gannett's Momslikeme does well, who the audience is, and what ads are sold for those sites. The idea would be to produce more content that advertisers would like.

But, Doctor said, it's a slippery slope because it removes editorial judgment. News publishers should use such formulas as a guideline, a way to understand the marketplace, Doctor said.

"They don't claim to be a journalism company -- they are a content company and that's fine. But it's not journalism," he said. "Journalism creates things that editors believe readers want and need to know about, some of which is commercially viable, some of which is not."

Using analytics or an algorithm as a guideline is something that's been taking shape for years. With the likes of The Huffington Post and the BBC experimenting with alternative headlines for stories and observing real-time traffic stats to determine story placement, content producers and editors are beginning to be more conscious of the demand from their audience. They rely less on instinct and more on hard numbers.

Demand Media realized that human instinct isn't always enough and that an algorithm can produce 4.9 times more revenue for each piece of content. This is why news organizations must use analytics to help shape their content plans, said Mark Briggs, CEO of Serra Media and author of Journalism 2.0.

"The ability to know what your audience is reading is powerful, and a news organization would be foolish to operate in the echo chamber of previous generations when top editors assumed what the audience wanted," Briggs said. That said, he doesn't believe in solely relying on analytics and said editorial judgment must be part of content decisions.

The shift toward demand-driven content illustrates how the Internet has shifted power to users, said Jay Rosen, a journalism professor at New York University. "Demand-driven means not editor-driven, after all," Rosen said.

And though the model doesn't seem applicable for news content yet, Weinroth said he thinks it is possible that shift will happen. News organizations could look at real-time trends on social sites like Twitter to guide their coverage, which can be used to drive traffic and support ads. 

Of course, a news organization that produces content solely based on its profitability breaks a trust with the audience that expects the news organization to provide information they need to know.

Jim Spanfeller, a former CEO of Forbes.com who plans to start a firm to help media with their digital businesses, said people come to a site because they have an established trust with that publisher and have certain expectations of how the editorial process works. Demand Media, on the other hand, shapes its content based on how users will respond.

"It's very different from news organizations that simply cover the news and then try to get their stuff out there more and have it be more visible through SEO," Spanfeller said.

Yet there's a difference, Rosen pointed out, between being guided by analytics and being obedient to analytics. "If people want to be informed by news providers they trust, the Demand Media way cannot be the way," Rosen said. "Neither the gods of click rates nor the priests of news should guide us.http://www.poynter.org/column.asp?id=31&aid=173972"

No comments: