What is a kelong? How to build a kelong?

How do companies like Demand Media and Mahalo determine what types of pages to build?

• I realize this is likely a complex answer but I'm looking for general guidelines as to how they go about determining which pages are the most valuable to build.

• Answer:

  Based on the Patent filed by Demand Media http://appft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&u=%2Fnetahtml%2FPTO%2Fsearch-adv.html&r=1&p=1&f=G&l=50&d=PG01&S1=20100153391.PGNR.&OS=dn%2F20100153391&RS=DN%2F20100153391 They use the following "method" for the ranking of keywords for profitability: 1. First they get the information regarding number of people searching for a search term i.e., Search Volume 2. Information regarding the amount advertisers are willing to pay for the search term i.e., Monetizability of the Keywords 3. Information regarding Click Through Rate of the search term i.e., the likelihood of the visitor clicking through to the site from the search engine i.e., Search CTR 4. Traffic Estimate of the Search Term i.e., the amount of the traffic expected based on different factors. 5. Most importantly, Longevity of the keyword is determined. Hence, "Halloween Costumes", a keyword that is expected to stay for a lot of time is given more importance (assuming other factors to be equal) than "Octopus Paul" which is less likely to have  longevity. Other than that, from the experience and observation, factors like Keyword Competition ( competition for a particular search term ), Hotness  ( relation to trending keywords on own properties ), Future popularity ( picking the probable keywords of future ) etc., also play a rule. There should also, probably, be a wonderful machine learning framework in the background that automates the process.