New AOL search analysis tool for marketers
I have finally finished working on the AOL search data tool that I hinted at in a previous post. I didn’t just want to throw out something that queries the AOL user searches and returns a flat list of results, although it is available.
The main goal for me to build this tool was to give marketers a tool that lets them research keywords or industry segments on a more aggregate level. I think it is very useful to do keyword research, understand user search behavior as well as doing industry research.
What’s the tool doing?
For a given search query, the tool queries against 36+ million records and retrieves the searches performed by AOL users. It then gives you the option of aggregating the result set a little bit. It’s nothing fancy, but you stumble over some interesting stuff when rolling up data.
Key Features
- Keywords are aggregated and the total number of queries & clicks are counted; along with it, the tool calculates the click though rate for every unique keyword.
- Click through rate by rank: this is an aggregation view by broken down by rank. Basically, for every position the tool counts the number of keywords and calculates a click through rate by position.
- Click through rate by page: similar to the CTR by rank aggregation/calculation, just that it is broken down by the page number. So you can compare what the aggregate click through rate was for keywords on the first page, versus keywords that appeared on the second or third page.
Please let me know if you would like to have additional functionality and I am happy to build it in.
Tip: You can exclude words from appearing in the result set by using a “-”. For example, if you wanted to look at all keywords that contain the term mortgage, but not calculator, you could search for “mortgage -calculator”
What’s next?
Admittedly, the tool is not lightning fast as it should be, depending on how broad your initial query is. I will start working on speeding it up, so stay tuned! If you are a DB guru and can help, please drop me a note.
Bookmark to:
