Alternative method of getting google keyword data

dxszzcylm

Insight for search shows you a graph with search volume, including seasonal dips, in a scale of 1/100.
They also provide the possibility to compare multiple keywords.
http://www.google.com/insights/search/#q=brown%2Cblue&cmpt=q

The data for insight are retrieved from these two pages,
google.com/insights/search/fetchComponent?q=brown%2Cblue&cmpt=q&content=1&lp= 1&cid=RISING_QUERIES_1_0
google.com/insights/search/overviewReport?q=brown%2Cblue&cmpt=q&content=1

The source of the last url is where the data for the graph comes from. Bash to give you a example;

1. $ curl www.google.com/insights/search/overviewReport?q=brown%2Cblue\&cmpt=q\&content=1 | grep "'.*,.*,.*\n'"

复制代码

will get you a long list of which i took a small snippet,

1. '1331856000,37.0,82.0\n' +
   
2. '1334448000,40.0,86.0\n' +
   
3. '1337126400,41.0,87.0\n' +

复制代码

The last line you can see 41.0 and 87.0 which is the search volume for "brown" and "blue".
Since the results are between 1 and 100 you can calculate how much traffic "blue" gets if you already know "brown" or vice versa.
So if you have one keyword of which traffic is known, you can use this method to estimate traffic of another keyword.469662131761

The data request url however blocks multiple requests a lot quicker then any other url you harvest data from.
Moderation should be used if you don't want all your proxies banned, you've been warned.

***UPDATE***
Small update, i forgot to include that the query url where you scrape your data is global.
Adding &geo=US gives query data for the states only forinstance.

Easy way to get query url variables for a specific nations and time ranges is to make a standard request in insight for search and then capture the browser requests to http://www.google.com/insights/search/overviewReport?
You can use wireshark or tamperdata for firefox, although i'm not to sure about the last one.