Additionally, paste this code immediately after the openingtag:
Let’s imagine that we are launching an online store for handmade mittens, and we want to create a web content that lets our potential clients easily find us. The first item on our to-do list could be to find some mittens-related keywords to use on our website.
But where do we start?
Google’s public data tool, more specifically Google Correlate, can give us meaningful insights about target keywords such as their correlations over time and across locations, and more.
If you are familiar with Google Trends, you can think of Google Correlate as Google Trends but in reverse. Google Trends tells you what is trending, i.e. patterns of search interests, while Google Correlate returns a list of search terms that share similar patterns with your inquiry. Without getting too much into statistics, Google Correlate uses Pearson Correlation Coefficient (r) to indicate the strength of the relationship between your inquiry and the list of resulting search terms. The r value is between 1 and -1; the closer it is to 1, the more positive the relationship is.
Because we want to open a handmade mitten online store, let’s enter “mittens” into Google Correlate and hit “Search Correlations” to explore other terms that share their search patterns with mittens. A quick glance at the result sees a list of the first ten search terms whose r values are the closest to 1. (see figure 1) We could also dive deeper into the result and conduct more analysis by downloading a CSV file, for instance, to look for terms that have negative correlation to our inquiry. (Check out Google Correlate’s tutorial page.)
An easy example of what a negative correlation might look like would be search patterns of swimsuits vs mittens. Additionally, we can observe where and when the search patterns are closely correlated to our inquiry and vice versa, by using the options to compare between US States, Weekly Time Series, Monthly Time Series, and shifting the time series by week or month.
Since mittens are winter accessories, clients are more likely starting to buy when the temperature is getting cooler. We could compare seasonal search patterns by weekly time series or monthly time series and take advantage of the shifting feature to move the time series by week or month to observe the change in correlation behaviors over time. Since most people do a lot of homework online before buying, this can guide us about their shopping interests and gauge when to roll out our ads and sale promotions. (See figure 1 for change in patterns and search terms with a 2 week-shift in time series.)
If we want to expand our inventory and client base, we can also utilize Google Correlate to gain more market information. See figure 1, with the box “Exclude terms containing mittens” checked, Knit hat, heated gloves, thinsulate, winter hats and fur hat appear at the top five. This implies that people were searching for these items as much, or as little, as they were searching for mittens over the same period of time. (Google Correlate displays terms that share similar search patterns, remember?) So if we want to add more products to our website, it might be worth considering selling knitted hats and other items as well.