Lately, I started asking myself, what is the formula for the way I choose keywords to optimize for on an existing page? To help people understand this concept better, I have written out an example. The actual equations are bolded, but this analysis should be read normally or you will miss important pieces of information.
I retrieve all of the search engine keyword ranking data, for example I rank in the 7th position for keyword Bake and in the 3rd position for keyword Able. I also retrieve my traffic data for each keyword (Able gets 100 hits a day, Bake gets 150 hits a day). How do I compare Bake to Able when they are both in different search engine ranking positions? I convert both keywords to what I estimate the #1 search engine position would get in traffic. I do this using data that AOL published on it’s user clickthrough data (Position #1 has 42% clickthrough rate, position #3 has 8% clickthrough rate and position #7 has a 3.3% clickthrough rate). So when I convert Able’s traffic from #3 position to #1 position, I get 42/8*100= 525. I do the same thing for Bake, moving it from position #7 to the #1 position: 42/3.3*150=1900.
Now that I know if I had the #1 position for Able, I would receive 525 hits a day and if I had the #1 position for Bake, I would receive 1900 hits a day. So I want to optimize for Bake obviously, right? Wrong. I still need to take into account conversion rate and the amount of competition for each keyword in order to determine which to optimize for. What if no users buy my product when they come to my page for Bake, but 5% of users buy my product when they visit my site for keyword Able?
Conversion rates per keyword should be data that you are tracking through your analytics package, whichever brand you use. For my example, I will give Able a conversion rate of 2% and Bake a conversion rate of 1%. I am going to multiply the #1 rank position traffic we figured out in the end of the second paragraph by the conversion rate for each to determine how many sales each keyword will receive. Sales for keyword Able= 525*.02 = 10.5 Sales per day. Sales for keyword Bake= 1900*.01 = 19 sales per day. Am I done yet? NO, QUIT ASKING! 😉 We still need to figure in competition. If it’s going to be 5 times as expensive to get the #1 ranking for Bake than Able, I don’t want Bake to be my primary keyword.
Competition is the hardest number to figure out. I’ll just give you the easiest way to estimate competition for a given keyword- use the Google Adwords keyword tool. Search Google for “keyword tool” and it is the #1 result. Type in “Bake” and hit enter, then type “Able”. Now hit go. The tool should give you a competition comparison (yes, this is for PPC, but easiest for newbies to use and somewhat accurate). I am going to use a sliding competition scale with 0 being the most competitive and 100 being the least competitive. For my example, Able has a competition rating of 70 and Bake has a competition rating of 40. Now I simply multiply my sales estimate for the #1 position we calculated in the previous paragraph by the competition rating. Able = 10.5 sales per day * 70 competition rating = 735. For Bake = 19 Sales per day * 40 competition rating = 760.
Thus, Bake is the keyword we should primarily optimize for, though Able is not far off. In a case like this I would recommend creating 2 pages, 1 for each topic.
IMPORTANT NOTE: The keyword/keyphrases mentioned above MUST be different words. For example, this would not work nearly as well for “Brass” and “Brass Balls” as it would for “Brass” and “Iron”.
If you found this to be a useful analysis, please let me know. I know it’s pretty dense, but it is much easier to understand if I could use a whiteboard to draw out the theory and examples.
Levi’s Question: it’s just theory right, as in, you have had success with it but it’s not been proven across the board, right?
Joel: it is just basic obvious stuff for any existing website
the only big hole is that maybe u have missed a keyword entirely
and that takes common sense to understand and research
for instance, I could be getting traffic for “teacher books” but I am completely missing out on the keyword “professor materials”