Share of Search is a very interesting KPI to track the strength of a brand in relation to other brands within the same category. In this article, we'll look closer into how you can measure Share of Search for any brand using common tools and free data sources.
If you're interested in learning more about Share of Search before you dive into calculating it, check out our article on the basics of Share of Search.
Share of Search can be calculated in a few different ways. The simplest and most reliable, as well as the method we're going to cover in this guide, is to use Google Trends. The process is fairly simple, but requires exporting and manipulating data from Google. This data is freely available from 2004 and onwards through Google Trends.
First, we need to get the raw data from Google Trends. Start by heading over to https://trends.google.com/trends and enter the brands you want to compare. Also make sure to choose the geography and time frame that you're interested in. This will give you a timeline chart over the trends data for the brands you've selected.
The data you see in Google Trends is the popularity of the search terms you've entered over time. Regardless of what terms you look up, you'll always see that one of them will have a single peak at 100 popularity and all other data points are relative to that. This means that the Google Trends data doesn't say anything about actual search volumes and also doesn't tell the full picture when it comes to the share of search of one term relative to the other at a given point in time. For example, you can see that during certain times the search volume increases for all brands, but the share of search will not as it is relative to the other brands.
In order to transform the Google Trends data to Share of Search, we'll need to do some manipulation. Start by clicking the download/export button at the top right corner of the timeline chart to download the data in CSV format.
If you need to compare more than 4 brands you have to make multiple exports. When doing so, ensure that you always have the biggest brand (the brand with the peak at 100) in all of you exports.
Time to import your raw data into your preferred spreadsheet software. In this guide, we're using Google Sheets, but the same principles applies in other tools as well.
When you've imported the data you should have a date column as well as a column with data for each brand. You can remove any rows above the title row (where your brand's names are).
You can then go ahead and add one more column for each of your brands, to the right of your current data. Call them "[Brand] SoS" or something similar. In the first cell of these new rows, we want to calculate the share of search for each brand. To do that, reference the original data for the brand and divide it by the sum of all data on that row.
Using the same logic, calculate the share of search for the first time period for the rest of the brand. You can confirm that you've done this correctly by checking that the sum of the new values are 1.
After doing this, you can apply the formula to all rows by hovering over a cell with the formula and double-clicking the small square in the bottom right corner.
Congratulations, you now have your share of search data! However, it's quite difficult to read and draw conclusions from. To change that, we're going to do some quick cleaning up and also display the data in a timeline chart.
First, select all the rows with your share of search data and convert it from decimal values to percentage.
Then, select the entire columns with your dates and brand's share of search data. Do this by holding Cmd/Ctrl and clicking the column headers. With the columns selected, insert a new chart (Insert > Chart). Change the chart type to line in the chart options panel if it isn't already. Make sure that the X-axis displays your months and that the Y-axis contains your recently added "[Brand] SoS" columns. You may have to add or remove values to the chart to accomplish this.
That's it, you've generated your first Share of Search report!
If the brands that you're comparing doesn't have really high search volumes, or if the search pattern is fragmented and the brands are very similar in share of search, your chart may be very difficult to read as the values fluctuate a lot over time. In order to properly visualize the data and to be able to draw any conclusions, you may need to apply a moving average to your data. By using moving average, we can transform each data point to be an average of the X data points before it.
First, decide on how many values you want to average out. The more fragmented your data is the higher value you'll need. Keep in mind though that the more you use averages the less details you will retain in your data set. We suggest starting somewhere between 2 and 5.
When you've decided, create another column for each of your brands. Call it "[Brand] MA [X]" or something similar. For example: "Toyota MA 3".
Then, for your first brand, go down one row further than the number of values you want to average. So if you chose 3, go down to the fourth empty cell. In this cell calculate the average of the first 3 values in your Share of Search column for the same brand.
You can then apply the same formula for all rows and for your other brands. You now have a moving average calculation of your share of search. This can be plotted in a timeline chart just like the previous example.
If you don't want to bother with exports and data manipulation you can of course use our free Share of Search generator instead. Just add the brands you want to compare in the form below and the generator will do the rest for you.