Start using Google Trends as part of our data analysis
Hi everyone, welcome back to my blog again! This time we will talk about Google Trends. Based on the explanation here, Google Trends provides access to a largely unfiltered sample of actual search requests made to Google. It’s anonymized (no one is personally identified), categorized (determining the topic for a search query), and aggregated (grouped). This allows us to display interest in a particular topic from around the globe or down to city-level geography.
My first time reading that my response was, “Woah, really? That was freaking cool!”
Now, I will try to share some basics about Google Trends so you can know how useful this feature is for our dataanalytics life.
Data Visualization Usage
Let’s say we are working at an e-commerce company, eBay. We would like to know if is there any correlation between the number of transactions made on our website in the year 2021 with the number of searches for a specific keyword, “ebay”. The data used in this example case is random. You can look into the data directly here.
As we can see above, the scale of the Y-axis we have above is 0 to 100. A value of 100 is the peak popularity for the term. A value of 50 means that the term is half as popular. A score of 0 means there was not enough data for this term. You can download the data from the graph above by using the Download feature, available at the right top corner of the graph.
Look we found here, we could see a similar pattern detected between the number of searches keyword “ebay” and the number of transactions made, we can calculate the correlation between those two variables by using the formula CORREL() in Google Spreadsheets. Based on the Merriam-Webster dictionary, the definition of correlation is a relationship that exists between phenomena or things or between mathematical or statistical variables which tend to vary, be associated, or occur together in a way not expected based on chance alone. In this case, linearly correlated, meaning they change together at a constant rate. The correlation value is 0.85 which means these 2 variables have a strong positive correlation. If let’s say the correlation value is -0.85 that means these 2 variables have a strong negative correlation.
With these findings, the most important question is: “How to find out if the correlation value above could mean causation?” We could create some hypotheses to be checked for, such as:
What is the main factor driving the steady decline in the number of transactions and number of search keywords “ebay”?
We could break down the variables by each constituent variable until we can find if there are any variables related. If we look back at the number of transactions, we can look into the main conversion funnel and finds out that the transaction was made because we got traffic. Of course, the result of Googling some keywords would result into generate traffic to our website.
That’s it, traffic! Maybe it has a strong correlation because both of these variables are connected by traffic.
After that, we can look to break down our traffic numbers to get into our website is there any pattern similar to our number of transactions or not, if yes, that could be “the causation” point we are looking for. :)
Interest by Region
Continuing our previous task, we could see below the first line graph, Google Trends gave us “Interest by Region” so we could know the search trend provided to us from which region or area. As we can see above, Germany is the highest contributor to our numbers, followed by the United Kingdom, Puerto Rico, Australia, the United States, and so on you can change the page to see other regions. We can also change the “region” into “city”. Very useful right, no need to do data tracking to know the location :).
Related Topics
The next information we can get is “Related topics”. Users searching for your keyword also searched for these topics. As default, the filter will be set to “Top”, but we can change the filter to “Rising”, which means related topics with the biggest increase in search frequency since the last period.
Related Queries
This part, Related Queries, means users searching for your term also searched for these queries. Using these matters, we can try to optimize our keywords when deciding to put more budget into our Google Ads by using these specific keywords. Ah yes, that feels good, right? :D
Another feature of Google Trends
You can find these cool things too, such as “Trending Searches” which tell us what is the top search trends, could be daily, real-time, and please choose the specific country you want to look into.
Or you could look into “Year in Search” which tell us the summary of search trend last year, you could also filter by country. The page would look like this:
As we could see from “Apa itu (What is)” it’s heavily related to COVID-19. From “Bagaimana cara (How to)” looks like most Google search users in Indonesia concentrate on how to get new, better, or additional income by being an online seller, creating a work application letter, and registering “Kartu Prakerja” and so on.
That’s all from me introducing Google Trends to you. If you already using this, that’s very great. I, as a data analytics practitioner, was helped by this feature from Google. If you never use Google Trends yet and feel like it could be great additional information for you to do your data analysis tasks, try it by yourself. I hope you guys find this post useful and big thanks to you guys if you read this post until finished. See you next time! :D
