One of the speakers at the WMW conference (sorry I forgot to note which one) briefly discussed keyword research and selection, using what he called the Goldilocks method. Simply put, it is this:
- You don't want keywords that are too hot! They are difficult to rank for, and are too expensive to purchase via PPC.
- You don't want keywords that are too cold! No one searches for them, or they are too general or not relevant.
- Obviously, you want keywords that are just right! These keywords will have the right mix of both relevancy and demand.
This analogy could be a great way of explaining keyword strategies to clients, so that they understand why trying to rank for "widgets" or "best little widgets in the whole wide world" might not be the best decision. The ones that are just right might be something like "automatic widgets" or "plastic widgets" or "widget planning", etc.
I usually run the "too hot" keyword through the Digitalpoint keyword suggestion tool and pick the group of keywords in the middle of the list as my starting point for optimization. (Of course, I weed out the ones that make no sense for my purposes). Usually, these will have the right mix of relevancy and demand. Once the site is ranking well for the majority of those terms, then I start focusing in on the too hot keywords. So, try the Goldilocks method with the next new site you optimize. You might find the strategy to be "just right".