Today I will cut right to the chase, keyword research is important. If you are planning on marketing your website and running some search engine marketing (or seo) then believe me, you must carry out some good quality keyword research to ensure that your marketing efforts are not in vein.
Thanks to an article in .Net magazine a couple of months ago, I learned a more speedy and robust method for carrying out quality keyword research that will leave you with data that means something and helps you to focus in on the keywords that are right for your business. So, I present to you a method of carrying out keyword research that is simple, straight forward, valuable and puts an end to any excuses against keyword research.
The Method
Any SEO expert worth their weight in Macbook Pros will know their way around Google's Keyword Tool, and if you do not, I covered it in basic detail last week in my PPC tools blog. The Google Keyword Tool is a fantastic weapon in your SEO arsenal but I know as well as anyone that the data it coughs up is no good without a good level of interpretation and understanding. This is what I want to introduce you to today. The method outlined below will leave you with an estimated number of clicks and sales a keyword will bring your website.
This method of keyword research requires you to begin by finding some keywords using the Google Keyword Tool and then pulling out the following piece of data; 'local monthly searches'. Let's say I am going to be optimising for dog toys and so I find the following keyword and data:
"Dog Toy" - 49,500 searches
Here we can see that the keyword 'dog toy' has an estimated local monthly search rate of 49,500. Take this data and keep it. The next thing you need to collect is 'competition rating'.
Competition Rating
To calculate your competition rating you need to type the keyword into your search engine (let's say Google) and then count how many of the first page results do not contain the keyword in whole. For every listing that does not contain the keyword in whole, count '0.1'.
When confronted with a long list of keywords to do this for it can be a little draining, so a good method is to use Firefox's 'find' feature to help you. Simply type your keyword into the 'find' area of Firefox and then click 'highlight all'. Then you will easily be able to count any search listings that have not been coloured pink by Firefox. This really does speed up the process. Another feature of Firefox (or other search engines) you should be using is 'private browsing' to ensure that the search results you see are un-biased.
For our keyword, 'dog toy', let's say that only two listings do not contain the entire keyword and so our competition score is '0.2'.
Sales Value
The final piece of data that you need to collect is 'sales value'. This is perhaps the most subjective piece of data as it requires you to say how many other things (other than the thing you are looking to advertise) can the keyword describe. For our keyword 'dog toy' you might say that this can describe both a 'toy for dogs' and a 'toy dog' meaning that there is one additional thing that the keyword could describe. For every additional thing you find, remove '0.1' from a total score at the start of '1.0'. Our remaining score for 'dog toy' is '0.9'.
The Winning Formula
Now comes the time to use all of the data you have collected and allow it to make some sense. There are two things that we can work out from the data: estimated visits and estimated sales / leads. To calculate the estimated visits you would receive from this keyword you use the following calculation:
Local Monthly Searches X Competition Rating X Sales Value X Click Through Rate for a Page One Google Listing = Estimate Visits
So for our example keyword, 'dog toy', the formula and result will be as follows:
49,500 X 0.2 X 0.9 X 0.15 (15% as a decimal - this is the avg. CTR for a page one listing on Google) = 1,337
So, immediately you can see that we have an estimated amount of visitors that we would receive if we were to achieve a first page listing on Google for the keyword, 'dog toy'.
The second thing to work out is estimated sales. In order to work this out you will need one additional piece of data that will usually involve delving into your website analytics package. You will need to find out your average conversion rate. If you struggle to find this, the global average sits somewhere around the 2% mark and so this can be your fallback.
The formula for estimated sales is as follows:
Estimated Clicks X Average Conversion Rate = Estimated Sales
And for our example, 'dog toy':
1,337 X 0.02 (2% as a decimal) = 27 sales
Giving Keyword Research Some Context
Immediately you can see that the research data now has context and this data will allow you to make informed decisions on which keyword to optimise for and which to ignore. Please note that although the data from Google is extremely useful, it is not always accurate to the exact number and so I usually use the data as a yard stick, rather than as an exact amount. Keep this in mind and your keyword research should now have the look of quality about it!