Mahalo: another content farm?.
In Calacanis.com , whose owner invested in Mahalo, it says that in this content site are four modes of learning:
1. Comprehensive search like Naver.com, Google or Bing (i.e. links, news, social, video, images, etc).
2. Content like Wikipedia or About.com
3. Question & Answer like Yahoo Answers, Vark or Quora
4. Educational Videos like Howcast or eHow.
It differentiates itself from algorithmic search engines like Google and Ask.com, as well as other directory sites like DMOZ and Yahoo! by tracking and building hand-crafted result sets for many of the currently popular search terms.
Looks good. But, is a content farm?.
yesterday Jason Calacanis, who recently pivoted his startup Mahalo to the content-farm model, called on the other content farms to pump out less content and better content. (Via Search Engine Land)
In an article of business insider about “mahalos-calacanis-to-other-content-farms-be-less-crappy”, “Calacanis pointed out that Demand Media puts out 5,100 pieces of content per day, AOL 1,700, Associated Content 1,500 and Mahalo 1,100 and would like to see those figures divided by ten.”
Mahalo, highly affected by Google Panda, is laying off 10 percent of its employees. In a message (via Center Networks) CEO Jason Calacanis and president Jason Rapp say the “changes have led to a significant dip in our traffic and revenue” and the company is therefore eliminating a “handful” of positions.
The company is also “pausing” its freelance content production and says it’s “determining how to best produce the high-quality educational material we aspire to in the long run.” That follows its idea of pivoting Mahalo from a “human powered” search/content factory to an education-centric enterprise.
Some people think that other important content farmer businesses were centered in Yahoo, for Associated Content and Answers, AOL for Seed, and newly public Demand Media.
Technically, a “from the ground up” approach was taken to develop a new content management system that allows content authors more control over their work. Additional functionality such as embedable content areas, infinite scrolling and “live answers” each increase the technical functionality of Mahalo’s system while moving it further away from their roots in Open Source Wiki.
Video Production in Mahalo
Jason’s announcement marked the release of a new design and some new front-end technology, but Mahalo 4 really started in June of 2010. In June, Mahalo pivoted from a revenue sharing model where authors were paid a percentage of on-page AdSense revenue to an hourly-rate model. At the same time, Mahalo launched a new form of content production called “Big Video”.
In the image below, Jason illustrates the start of the “Big Video” project along with the increasing size of Mahalo’s video library.
These videos earn money both by being included on Mahalo’s own SERP’s (Search Engine Result Pages), on YouTube and a wide base of other smaller websites through flash inclusion.
Mahalo exists to earn money. There is no other reason. Mahalo’s videos earn the company money in two different obvious ways. First, Mahalo relies on Google’s partner program. In short, the YouTube partner program shares part of the revenue that Google receives on the ads they display when a video is playing.
Next, Mahalo earns money on the videos they create by placing them on their own SERP’s. Along with written content, on site videos increase retention rates which in turn, provide more time to encourage users to click on the AdSense ads placed on every Mahalo page.
The third and less obvious way that Mahalo earns money on videos is through traffic generation. Online traffic is a commodity as valuable as oil is in the real world. Not only are Mahalo’s YouTube videos embedable anywhere online but there are various other tactics that large video collections can use to increase traffic.
Mahalo has a good number of pages that truly define good content. Mahalo has been working hard to improve their content because Google (and other search engines) are starting to identify content farms that degrade long tail searches.
In a very interesting article, Robisit (mahalo version 4) says that “It turned out that the model Jason described to Fred was first to hire a small army of employees, then to contract offshore firms to create SERP’s. When those things didn’t produce the volume or quality of articles required to make money, Mahalo devised a couple of different ways to motivate remote, unknown North American labor to become article authors. Motivation always centred around money – Mahalo SERP authors received between $2 and $10 per article.
In each of those scenarios, the focus was never on quality. How could it be? Even if you were an excellent author with specific experience in what you were writing, Mahalo’s model wasn’t to pay you a reasonable rate for reasonable work.”
The article continues: “Mahalo’s business centered around people. I think that Jason’s failure to build Mahalo into a sustainable business was a result his inability to foster and grow the communities of contributors that were required.”