Last week our CTO gave a presentation on Correlation vs Causation. It gave us pretty good idea that a correlation between variables does not necessarily mean that one variable is the cause in the change of the other variable. You can find the Presentation here.
I was really fascinated by the PPT and while I was going through a book which was somewhat on the same lines of the ppt, I came across Google Trends.
Google trends is a search tool which allows a user to see how often a particular keyword has been searched online in a specified period of time. Not just that, it allows you to see how a particular keyword may be correlated to another keyword. You can see the trends along different countries and regions.
Now the data from the Google Trends seems a bit confusing at first, the output may seem like the frequency of search on google of that particular keyword. The higher the value, the more number of times it has been searched? Correct? Wrong.
It is actually the ratio of that keyword search to the total number of searches.
Let me explain that with an example, an example with my favorite food item, “Butter Chicken”.
Which Country do you think is on the top of Google Trends when it comes to searching Butter chicken? Seems like India? But it is not. India in fact has the 11th position. New Zealand is number 1 and Australia is number 2.
This just means that though India is obviously being searched a lot in India, it does not have a higher ration compared to other things that are being searched.
If you want to check the search volume you can use the Google Keyword Planner.
Now this may seem as a fun tool, but it has a lot of useful applications too.
Google searches are honest searches, people are completely honest over the internet.
Google Trends can help you find the real time marketing opportunities, you can see how a particular topic of interest has spiked over the period of time and in different regions.
It can help you compare your brand compared to your competitors in different regions and see the interest of the people.
Given below is the comparison of Dominos and Pizza hut in different countries
You can also predict election results through this platform, before a major election you can check out which is the more liked candidate in different regions. Now one may argue that a person might search for a candidate if they hate them too. But it has been found that when searching for 2 candidates like Trump and Hilary, people would put the first name of their liked candidate first like Trump-Hilary election.
Google correlate is a similar tool, which can help you find similar keywords. Say I am running a blog on Business Intelligence, google correlate gives me correlated keywords as given below
So I should write more blogs about sql and database because people are interested in them as well.