Google Correlate is a new experimental tool launched by Google on the 25th of May to determine search patterns that match up with real-world trends. It may be thought of as the reverse of Google Trends. When youâ€™re using Google Trends, you have to type a query to get a series of the frequency of that query. In the case of Google Correlate, you have to enter a data series to get queries whose data series follow a similar pattern. This is currently however only a Google Labs project.
Using the new tool you can trace search terms that vary in popularity over a period of time. To find search terms that have a similar pattern of activity as your own time series you can enter your query into the search box.
To identify search terms that have a similar pattern of activity across the US, just as your own US states dataset, you can enter your data into the search box.
How it Works
When you enter a data set, for instance a time series, Google Correlate works out the Pearson Correlation Coefficient (denoted by r and values range from r=-1.0 to r=+1.0) between your time series and the frequency time series for all queries in Googleâ€™s database. The results returned are the ones having the highest correlation coefficient (a value closest to r=1.0).
Hereâ€™s a look at the benefits of this new tool:
- Results can be surfaced in real time
- You do not need any prior knowledge to make a good choice of queries that match the phenomenon. With only a spatial or temporal pattern of interest, the queries that are most similar to the entered data are returned.
- Can be helpful for search engine optimization keyword research. By entering the important brands and generic keywords that are powerful in your niche, you can determine what searchers in your niche are searching for.
- Though Google returns only positive correlations, you can see the negative correlations also if you want to. For that before uploading it to Google Correlate just multiply your input data by -1 in your spreadsheet program.
- Privacy â€“ None of the information in this tool can be linked to a specific person. The data incorporates no information concerning the identity, specific location, or IP address of the user. Apart from this, Google would anonymize any original web search logs that are more than nine months old.
- Filtering â€“ Queries which may not be interesting are filtered. This includes rare or misspelled queries, queries of a correlation value that is less that r=0.6, and pornographic queries.
If we can make the most of Google Correlate, it would surely help the cause of internet marketing.