The first point regarding keyword research and selection is that you should first select the "desired" search terms, those that would be most valuable because they reach the target audience best. This is important because you're building a website for humans, not search engines.

Selecting your most desirable search terms has nothing to do with the level of competition. It's all about reaching the target market. The process I recommend is the relevance evaluation and "weighted popularity" calculation; our spreadsheet is available for download from the tools area on our site.

If people are searching with certain words and phrases, they expect to find content around those terms. Even if you believe that the SEO competition is too great for a certain generic phrase, you still shouldn't build a site without that content.

Keyword research should tell you what people are looking for; a good site will provide it, regardless of SEO considerations.

The second point is that competition may dictate how you target the desired search terms. Some will be reachable via SEO with an acceptable level of effort; some will be reachable via PPC at an acceptable cost; some will be out of reach (vs. the budget) either way.

Just because you can't target the generic phrase, that doesn't mean it's not important for your SEO strategy. People will use that generic search term as part of longer searches, so it's important to identify the "modifiers" that people will use, and incorporate those in the content. This allows your "for humans" content to attract search engine referrals.

As an example, we worked recently with a site that has information on real estate licensing & training programs. All of the generic terms were too competitive for their budget, but with rich content their referrals for terms that include "real estate license" exceed the projected traffic for a top-10 ranking on that generic term.

With the more refined competition metrics (intitle+inanchor) there is no "rule" as to what makes a search term too competitive. This metric gives you an indication of how many pages have been optimized for any given phrase. I would say that if the "intitle+inanchor" count is very close to the "intitle" count, it usually indicates a higher level of intentional competition. We like to look at the link-popularity numbers from Alexa (# of unique web hosts linking to a site) as well.

One other point about competition: In my opinion, the best measurement of competition is pay-per-click. If people are willing to pay $5.00 per click with Overture, they're also willing to spend that much to get traffic from organic listings. This varies with the seasons on some terms, but so does SEO competition, now that link buying is such a big part of the landscape and the focus of a site's external profile can be changed from month to month.

Finally...in most cases, a really disciplined approach to assessing relevance will lead to good keyword selection for SEO, and competition becomes less of an issue. Measuring the current competition in any way isn't the same thing as considering the long-term competition anyway.

The competition is what it is; it will continually increase for all search terms. The main thing you can do about this is to improve the effectiveness of your link-building efforts, and there are many ways to make link building a profitable activity in itself.

If you teach people to build links profitably, then they have a huge long-term advantage vs. folks who are simply buying text links from link brokers. If my links pay me every month, and my competitors' links send them an invoice every month, who is going to win that race?

Dan still has room in his Advanced SEO class. He's also running an "SEO Your Site" workshop which is intended for folks who have one website they want to work on. The workshop also includes a complete video set from the Advanced SEO class. You can sign up for the Advanced SEO class here.

Discuss this article in the Small Business Ideas forum.