The folks at Xendant did an excellent analysis of the best Adsense keywords. Their post/article makes one simple point. It’s not just about the cost per click, but its about the cost per click times the number of clicks you can expect.
They present two tables: one that sorts keywords by highest CPC, and the other is sorted by CPC * number of expected clicks. Boy do these two tables tell a different story.
Of course, there are other things to consider. Many people have been aware for a long time where to make the most money, so the most profitable keywords outlined in the Xendant tables are also the most competitive. You don’t want to leap into promoting an insurance, travel, or education site on the web lightly. The competition is very, very tough, and it’s backed by money.
So you do need to pick your spots. You need to figure out how a table like this relates to unique value add that you can bring to the table. Fortunately, if you bring a significant amount of unique value add, you should be in good shape. Much of your competition is crap, and will eventually fall by the wayside.
However, you need to anticipate that this will take a long time. If you are building a site from scratch and want to build an SEO empire for one of these categories, it’s a minimum of 6 months after launch before you get to interesting traffic, unless you somehow come up with a killer idea. Patience, intense focus on the long term goal, and determination to stay on course will all be needed.