SERP Competition Analysis

[Solarius]

Hi! I was wondering how you measure the competition or optimization level of a SERP for a given search phrase. Do you have any way of quantifying this or do you use a more intuitive approach?

Since I am new here, it might be that this has been discussed earlier, and in such case I apologize in advance. I have, however, looked around in the forum and haven't been able to find anything that addresses this specifically.

After reading about SEO for a week and a half, here are some of the ideas I have picked up:

1) KEI/KDSI
On numerous websites, and in different software programs I think, they recommend different variations of this concept; i.e. dividing the search volume for your search phrase on the number of URLs in the returned SERPs. The result of this division is then supposed to tell you how profitable the keywords/search phrase is.

The search volume is obviously of crucial importance, but I fail to see how the number of URLs in the SERPs can tell us something about how likely it is that we will succeed or not. According to research by West Point (1997) and eMarketer (2001), 77% of the users do not go beyond page #2 of the SERPs. This means that SERP #3, #4,... should be irrelevant for determining whether one will succeed or not.

If Google reports that there are 3.500.000 URLs that contain your search phrase, it might be easier to get a top placement there if none of the pages are optimized for the search phrase than in a set of SERPs containing only 350 URLs, but where the top pages are highly optimized.

The question, as far as I see, has to have focus on the SERP one wants to be in; i.e. how highly optimized are the URLs in SERP #1.

2) Backlinks
Another method to measure the competition is to count the number of backlinks. This is, in my opinion, a much better approach. Based on research I've read, the number #1 position at Google has on average more backlinks than any other position in the SERP, just like the average #2 has more than nr #3.

However, this is only an average. In a concrete SERP it might be different, since the value of a backlink can vary very much. The value of a backlink, is decided by the anchor text and the amount of PR it carries (which is determined by the PR of the page where the link sits, and the number of other outbound links it has to share the PR with).

When considering whether it is worthwhile to target a SERP for a search phrase, it could be very valuable to be able to determine the optimization level for this phrase more precisely than counting links. Let's say that the top position has 2300 backlinks, but that only 300 of these are relevant for the search term, then maybe instead of trying to secure more than 2300 backlinks for this phrase you could aim for a lower figure which would make the economics of the project completely different (and maybe use the other backlinks to target another search phrase)? But how would you go about to do this? You can't check each link manually. When you start working on acquiring backlinks with the relevant search terms as anchor text, you will discover this by checking your ranking, but it would have been interesting to have an indication in advance.

For me all of the above is theory, so I would be very interested in hearing what you think about this. How do you measure the competition of a SERP for your search phrase before deciding what steps are necessary and whether it is a worthwhile target?

[ferret77]

go to overture and see how much people are paying to be in the number 1 spot

its usually a good measure

[compar]

Well, those are interesting questions. Oh if it was only that simple.

To start with most people agree with you and think the KEI as calcuated by Wordtracker is virtually meaningless. Your competition for any search term isn't the number of pages returned for the search. Your competition is the 10 or 20 top pages. Those are the only pages you are worried about beating.

Now to your second point. There is no measure of "optimization". There is only a measure of success -- are you in the top 20 or not? There are a 100s of thing that Googles algorithms evaluate and Google won't tell us what they are.

On this forum in particular we feel that backlinks are a very crucial and important element, but nobody knows exactly how the value of backlink is measured. PageRank is only a part of it and in my opinion a part of declining importance. We think that the recent changes in Google reflect a new way of measuring the relevance of a backlink, but nobody know that for sure or how it is being done.

So there is no universal tool for measuring the relative degree of optimization. Just look at the top ten pages for the keyword phrase of your choice. Do your own assessment and determine whether you could do a better job.

It depends on the particular search term but often you will find that even site listed as high as #5 have no obvious signs of optimization and should be easily surpassed.

[Solarius]

Thank you very much for your reply, Compar. After reading it, I went out and read every article I could find on this site. Now it seems like I should visit Brian's new site as well.

Anyway, I think you are right that it is difficult to find an easy way to quantify the competition level of a particular SERP. Therefore, what you say about developing one's intuition has to be the way to go. An intuition has to be based on some factors though, and backlinks and PR appears to be two good starting points. I was wondering whether there is a third factor that also might provide some help, namely the number of pages the sites in the SERP have had indexed by the SE. What do you (and others) think of this?

When looking at different search terms that will be relevant for my site, I discovered that what you wrote in your last paragraph is true. Site #1 - #3 had between 2500-20000 backlinks (Alltheweb) and 3000-10000 pages indexed (by Google). Nr. #4 - #10 had on average 200 backlinks and only 400 pages indexed.

Will using the backlinks from Alltheweb and the number of pages indexed by Google like this, be similar to comparing apples and oranges? I tried to find the backlinks by using Google but the numbers seemed to be just a selection, so I decided to use Alltheweb. ...and the number of backlinks in Alltheweb seems to fit rather well with the ranking in Google. Anyone with any thoughts on this?