jump to navigation

Demand Media: Search Spam or the Future of Content? January 31, 2011

Posted by David Card in Uncategorized.
Tags: , ,
add a comment

Demand Media’s successful IPO last week didn’t quite dispel the air of controversy surrounding the company. Demand has never been profitable, it and its content farm brethren are blamed for diluting search results and its content is damned as lightweight fluff written for near-slave wages.

But the controversy is overstated. Though Demand’s business model appears a little shaky, companies in online media should monitor it closely for lessons in content creation efficiency and targeting, and for potential partnership opportunities.

How It Works

Demand produces low-cost content to order based on its analysis of search trends, advertising rates, competitive content and its long-term value in terms of ad sales over time. It suggests topics based on that algorithmic analysis to a network of 13,000 freelancers. Those freelancers then write articles or make videos that Demand shows on its own highly search-optimized sites — eHow, Answerbag, Livestrong.com and Trails.com — or syndicates to partners like the San Francisco Chronicle and the NFL. It pays its freelancers $15 for short articles.

But this cheap, clinical approach to content creation inevitably offends journalists, and produces a lot of how-to articles of varying quality.

Limits to the Model

Demand Media content is at least mildly useful to the masses. Otherwise, it wouldn’t generate CPC revenues or enough clicks and links to influence Google results. That content is not hugely valuable, but it’s not as worthless as this this story, which compares Demand’s revenue per user to that of Google, implies — Google, after all, is one of the most profitable media companies in history. In contrast, Demand’s dollars per thousand page views on its own sites is showing modest growth, indicating the company is starting to move beyond low-CPM text ads into higher-value advertising.

In the most recent nine month period, Demand depended on Google for 28 percent of it revenues. It uses Google’s ad networks and has a revenue-sharing agreement for video content shown on YouTube. Two-thirds of the page views for its biggest site, eHow, come from Google, and search drives 40 percent of the views across its network.

But that dependence is one way. Google takes about a 15 percent cut of the ad revenues its network generates for partners (compare Traffic Acquisition Costs versus network revenues). So the $50 million Demand got from Google over nine months produced peanuts for the search giant. Any conflict of interest implied by Google making money from content farms just isn’t worth it, compared with the potential damage to Google’s reputation. Nonetheless, Google recently responded to chargers that content farms and spam were affecting search results quality. And Demand’s attitude that Google’s response had nothing to do with it isn’t entirely convincing. Demand’s Google dependency is a little scary.

How It Could Evolve

Yahoo acquired Associated Content, a content farm similar to Demand, and AOL has a rival freelance network called Seed, run by ex -New York Timesman, Saul Hansell. AOL’s Patch hyperlocal content network isn’t completely different from a content farm — and has search engine watchers equally concerned. Content farms are an evolution of the Times’ About.com property, and the Times considering investing in Demand. So clearly, Demand’s content farm model is attracting attention. Here’s how Demand might evolve it further:

  • Verticals. Some Demand properties align celebrity brands with topics (Lance Armstrong-health, Tyra Banks-beauty), but the company hasn’t invested much in higher-quality content for those sites, or in other verticals like technology, food, or older audiences that could command premium advertising.
  • Sponsored Content. With just a little polishing, Demand could apply its model to advertorial-like sponsored content and combine that with the social media tools it acquired via Pluck for advertisers.
  • Q&A and Content. Blending community-answered questions a la Quora with a content farm seems a natural fit – Mahalo may be steering this direction with video already.

Related Research: Google Needs to Fix its Spam Problem, Even if it Hurts

Question of the week

What can companies learn from Demand Media?

Google’s Chrome OS: Dead Before Arrival? December 13, 2010

Posted by David Card in Uncategorized.
Tags: , , , , , , , , ,
1 comment so far

Last week Google showed off its progress on Chrome OS, introducing an apps store in support of it and offering up a pre-release hardware trial program (real machines won’t ship until the middle of next year). But it’s likely all for naught. Google CEO Eric Schmidt’s objective of making Chrome OS a “viable third choice” in operating systems looks doomed.

The Problem With Chrome

Right now, the hot trends in technology are social, real-time, mobile and cloud computing. Chrome OS is only optimized for one of them — its machines are true cloud clients. Schmidt even evoked the old Network Computer vision. Chrome OS computers will be highly dependent on the cloud for applications and minimally functional when disconnected. They’ll have cellular modems, but it’s not clear that existing networks can handle the network traffic demands of a cloud-centric client. Meanwhile, there’s nothing in Chrome OS or its user interface that accommodates social media or real-time information feeds.

Chrome OS also suffers from awkward positioning, both externally, to developers and potential customers, and internally within Google’s own product line-up. While it’s true that PCs serve both companies and consumers, the value of the Network Computer premise appeals only to enterprise IT managers. Its manageability and simplified functionality play best in applications like airline reservations, point of sale terminals and ATMs, or in limited-application mobile devices used in shipping and store inventory management. Yet at least for now, app stores are purely consumer offerings. The apps Google showed last week all came from media companies (New York Times, NPR, Sports Illustrated), Electronic Arts and Amazon.

Opportunities for a New OS-Based Platform

Meanwhile, Google itself says Android will be its primary tablet operating system. In fact, Google aims Android at most of the best opportunities to establish new or alternative operating systems. I’d argue that there are three product categories where Google could try to establish a new OS platform, either with Android or Chrome OS:

1) Mobile phones. Google has Android.

2) Tablets. Gesture-based tablets have all the momentum over netbooks, and are also Android targets.

3) Single-function Internet devices based on a customizable OS kernel. This category also includes The Internet of Things, with its connected gas pumps and refrigerators, that may be better served by something that looks more like a sensor than an OS. But other single-function devices include, in order of current volume:

  • TV set-top boxes. Google TV is based on Android, with a Chrome browser-based UI that shows its limitations.
  • Game consoles. The “compatible console” concept with hardware running a third-party OS has not worked so far (e.g., Sega Dreamcast with Windows CE, 3DO.)
  • E-books. Barnes & Noble’s Nook Color already runs Android.
  • Widget machines. Chumby has set the standard for “Internet viewers.” The Sony Dash is built off Chumby.
  • Dedicated social devices. Just like Chrome OS, Jolicloud uses HTML5 and a Chrome browser on top of Ubuntu Linux, but it uses Facebook Connect heavily. Handheld social devices like the TweetPeek and Sony Mylo seem to be going nowhere, while others include phones (Sidekick, Motoblur, Kin).

Those options don’t look promising for Chrome OS. They’re either tiny markets today,  already staked out by Android, or, in the case of game consoles, brutally competitive and unproven for third-party OSes. I suppose Google could fight it out with PC operating systems based on price. A Microsoft OS makes up 10 to 20 percent of cost of netbook, but Schmidt says Chrome OS devices should cost $300 to $400, about the same as netbooks. An Intel processor, solid-state storage, and an integrated cellular modem raise costs. I’m tempted to agree with Om, and suggest Google dump Chrome OS and put its muscle behind Android.

Question of the week

Should Google dump Chrome OS and put its muscle behind Android?

Real-Time Advertising: How to Get in Early October 18, 2010

Posted by David Card in Uncategorized.
Tags: , , , , , , , , ,
add a comment

Even if you don’t believe real-time feeds will become the dominant content consumption paradigm, it’s clear they’re a growing force. Consumer-paid access to real-time feeds is largely constrained to paid mobile apps today, so advertising would appear to be the immediate payoff. With that in mind, let’s look at how social media companies can best cash in.

I haven’t come across a forecast for real-time advertising spending, but it’s a nascent market that’s fairly concentrated: Facebook and Twitter represent the largest audiences. Market researcher eMarketer projects Facebook will collect $1.3 billion in ad revenues globally in 2010; presumably, most of that will be spent on Facebook’s news feed. Twitter is only just beginning to embrace advertising. But clearly, we’re talking about a business that will be measured in billions rather than millions of dollars.

The current audience concentration — and the resulting ad dollars — could diffuse. Already, a lot of tweets get viewed on third-party Twitter clients, as well as on Facebook. Somewhat similarly, Facebook is syndicating its content through initiatives like Facebook Connect and Instant Personalization, as well as arrangements that allow companies like Skype to show Facebook users’ updates and presence within its own application. So it might not be just Facebook and Twitter who can cash in on those audiences.

Could Real-Time Ad Networks Jump-Start Spending?

Advertisers demand a certain scale of audience before they start spending big money. As with other media — social or otherwise — ad networks can alleviate audience fragmentation, giving advertisers access to eyeballs across a number of sites or apps. The big ad networks from AOL, Google, Microsoft or Yahoo aren’t doing anything in the real-time space. Meanwhile, a handful of startups have emerged. That includes 140 Proof and OneRiot, who sell inventory on Twitter clients and apps, as well as Tweetup, which also makes its own destination site. Ad.ly will construct celebrity-sponsored updates and insert them in Twitter and Facebook streams.

I spent some time with OneRiot this week; its experiences are good indicators of the state of the real-time ad marketplace:

  • Tapping test budgets. OneRiot’s business is divided evenly between publishers (New York Times, ESPN, Guardian) who are promoting its stories or marketing their apps in feeds and more traditional marketers like Zappos and Stella Artois. OneRiot is getting part of the test budget of bigger campaigns, so advertisers are only spending tens of thousands of dollars with it. The company can charge 12 to 25 cents for click-throughs, or $2 to $3 CPMs.
  • Relatively simple targeting. OneRiot usually sells an audience type rather than target by demographic or content context. Its analysis shows that Twitter client users are a highly engaged audience; when they click through to a story or site, they’re likely to hang around twice as long, generating 7 or 8 pageviews.
  • Ad format experiments. OneRiot serves up text ads that look like search engine marketing, but its architecture can handle banners and richer formats. It says some advertisers have experimented with dynamic content that is contextually related and inserted into the text creative.

Ad Network Realities

Right now, the ad networks in real-time are ahead of most of the feed sites in sophistication, and could help move the market forward. But in most media markets, it’s the company with the eyeballs that commands the vast majority of ad spending. Not long ago, observers who probably over-interpreted Google’s success thought online ad networks could reverse this. But that hasn’t turned out to be the case.

Publishers and other content companies like to hold onto the best ad inventory and sell it directly to their best advertisers and ad agency clients. That leaves low-priced remnant inventory for the networks. NBC dropped Google’s TV ad network recently; Microsoft is shutting down its in-game ad network because its biggest customer, Electronic Arts, pulled the business in-house.

As with other media, the real-time ad network ecosystem will have to deliver targeting and measurement to capture advertiser spending. Companies like Klout and Gravity may help marketers identify influencer audiences. Kantar Media, a unit of ad agency holding company WPP, tracks offers competitive intelligence on ad networks, but hasn’t aimed at the real-time space yet.

Related Research: Social Media in the Enterprise

Question of the week

Could ad networks accelerate real-time advertising?