Don't compare apples with oranges
Posted: Sun Jan 05, 2025 10:41 am
Now, you can probably start to see why it’s a huge tool for content marketers. It provides free, highly customizable, and authoritative data, and people believe it because it comes from Google.
But let's take a closer look at the information we're receiving. You might notice something unique on the Y axis, which brings us to the first rule.
#1.
In all Trends charts, 100 represents the highest search volume on the chart; in this case, it's the day that "Old Town Road" was searched for the most.
This is the case for this chart and all Trend charts, because all charts will, at some point, have their highest point. All other values on the chart are relative to that 100 value.
In other words, each number along the line is a percentage of the time frame at its peak. So, since the number is around 50 at the bottom of the chart, we can assume that on the last day of the interval, about half of what was searched for was done at its peak.
However, the main problem we see is that some content creators try to compare terms in separate charts without realizing that they are completely independent of each other.
If you were to search for "Baby Shark," you'd still see a dot at 100 somewhere (as you would for any term). It wouldn't be fair to compare that 100 to the 100 for "Old Town Road," because we have no idea what the actual search volumes are.
Rely on Google Trends data - SEMrush
To do this properly, you'd need to use Trends' handy "Compare" feature, where you can see that "Old Town Road" had almost twice the peak search volume, even though "Baby Shark" has been around longer.
Rely on Google Trends data - SEMrush
Since search interest is relative, we can't technically say, "It had a search volume of 78," because in reality, the value 78 is representing 78% of the maximum search volume.
Looking back at the “Old Town Road”/“Baby Shark” battle for my young nephew’s attention, you may have noticed something: while “Old Town Road” had the higher high and has cambodia phone number database remained above “Baby Shark” since its release, the “Average” bar chart on the far left shows “Baby Shark’s” bar significantly higher — 26 to 15, to be exact (which, as discussed in Point 1, only means 26% and 15% of the chart’s highest volume, which was “Old Town Road” in April 2019).
That would lead you to believe that “Baby Shark” was, on average, ~80% more searched than “Old Town Road.” Of course, that’s misleading, given the time frame we selected, where “Old Town Road” didn’t exist. If you’re going to compare things, they always have to have had the same chance of getting search volume during the selected time frame.
How to apply the rule
Content creators must be responsible and accountable for the numbers and information they present. That means they have to do ("do do do") the necessary due diligence to understand their data, what it represents, and that the claims they make are fair.
One test I like to use is: If I were an editor who didn’t want to cover this story, what would I say? This approach allows you to be critical of your data and allows you to preemptively address problems before they arise.
But let's take a closer look at the information we're receiving. You might notice something unique on the Y axis, which brings us to the first rule.
#1.
In all Trends charts, 100 represents the highest search volume on the chart; in this case, it's the day that "Old Town Road" was searched for the most.
This is the case for this chart and all Trend charts, because all charts will, at some point, have their highest point. All other values on the chart are relative to that 100 value.
In other words, each number along the line is a percentage of the time frame at its peak. So, since the number is around 50 at the bottom of the chart, we can assume that on the last day of the interval, about half of what was searched for was done at its peak.
However, the main problem we see is that some content creators try to compare terms in separate charts without realizing that they are completely independent of each other.
If you were to search for "Baby Shark," you'd still see a dot at 100 somewhere (as you would for any term). It wouldn't be fair to compare that 100 to the 100 for "Old Town Road," because we have no idea what the actual search volumes are.
Rely on Google Trends data - SEMrush
To do this properly, you'd need to use Trends' handy "Compare" feature, where you can see that "Old Town Road" had almost twice the peak search volume, even though "Baby Shark" has been around longer.
Rely on Google Trends data - SEMrush
Since search interest is relative, we can't technically say, "It had a search volume of 78," because in reality, the value 78 is representing 78% of the maximum search volume.
Looking back at the “Old Town Road”/“Baby Shark” battle for my young nephew’s attention, you may have noticed something: while “Old Town Road” had the higher high and has cambodia phone number database remained above “Baby Shark” since its release, the “Average” bar chart on the far left shows “Baby Shark’s” bar significantly higher — 26 to 15, to be exact (which, as discussed in Point 1, only means 26% and 15% of the chart’s highest volume, which was “Old Town Road” in April 2019).
That would lead you to believe that “Baby Shark” was, on average, ~80% more searched than “Old Town Road.” Of course, that’s misleading, given the time frame we selected, where “Old Town Road” didn’t exist. If you’re going to compare things, they always have to have had the same chance of getting search volume during the selected time frame.
How to apply the rule
Content creators must be responsible and accountable for the numbers and information they present. That means they have to do ("do do do") the necessary due diligence to understand their data, what it represents, and that the claims they make are fair.
One test I like to use is: If I were an editor who didn’t want to cover this story, what would I say? This approach allows you to be critical of your data and allows you to preemptively address problems before they arise.