April 16, 2019

How to Interpret Google Trends Data

This is most applicable to the many ways the broader media misuses data in reporting, but I think a lot of marketers and generally curious individuals will benefit from reading this, hence I’m posting it here.
Posted in: Analytics
President & Lemonhead Evangelist

View More

Everyone is talking about this black hole picture!

This is most applicable to the many ways the broader media misuses data in reporting, but I think a lot of marketers and generally curious individuals will benefit from reading this, hence I’m posting it here.

One of my many pet peeves is when I see a reference to “Google search data” in an article, tweet, whatever. Almost universally, those references are referring to Google Trends.

Trends is one of the more accessible tools Google offers. It uses simple, yet appealing visualizations. It’s free and requires no account. It’s pretty darn close to real-time. And it purports to give you a gauge of the popularity of something in Google search over a pretty long period of time. What’s the problem?

It’s a big problem. Way too often, folks referencing Google Trends will take the 0-100 scale to be an absolute unit of measure. In the image above showing interest of the term “black hole”, that would mean that for a recent moment in time, “black hole” was basically the most popular thing on the internet. And there’s plenty of confirmation bias to back it up! You no doubt saw tons of news about the black hole photo, all your social media was flooded with it. It was a big deal!

But the most popular thing online? Far from it.

Let’s use the compare feature to get an idea of just how popular this event was.

What happened???

Each of the last five Masters tournaments was way more popular than the black hole. Each was easily 2x on this 0-100 scale. But how can that be? Before, the black hole was a 100, now it’s like 30-35.

That’s because this is not an absolute scale. It’s relative. When looking at a single topic/term, the line represents the popularity of a term over time relative to the lowest and highest points of all time interest in that term. If something hits 100, it just means there has never before (on Google) been more people interested in that topic. Not that it’s the most searched/viewed thing period.

When we look at the Masters, we see the last two tournaments hit peak popularity for the history of the Masters on Google. And when comparing it to the black hole, we again see relative to the Masters, there’s a lot less interest in the black hole.

The key is to look at the name of the product. It’s Google Trends. It’s not a measure of volume of searches. They have separate tools for that, which aren’t as accessible to the general public and also (often) not as close to real time. This is a measure of velocity or change over time.

So if you were reporting on the black hole, for instance, it would be incorrect to make any sort of claim that the black hole was one of the most searched topics on Google. It would be correct to say that more people were interested in black holes than at any point in recent history (with the assumption that “recent” refers to the modern internet age).

The more appropriate way to use this tool is for gauging whether longer-term interest in something is generally increasing or decreasing. Here’s a few examples. I’m using photo captions to make very general inferences about the data – to give you a little better idea how to make use of this tool.

Looks like Chevy has done a great job rebounding their brand after the financial crisis
Before saying people aren’t interested in this topic, I’d want to lay out what happened in 2006-07 and explain how that relates to more recent years. Basically, spikes in interest can be misleading.
You could make a case that Apple is flattening and the timing of this leveling off seems to be ahead of their sales slowing by a bit.
I guess Justin’s mix of Despacito wasn’t enough to keep his career going.

This is a pretty simplified overview of how to use Google Trends, but hopefully leaves some of you in a much better position to properly relay the data in your writing, conversations, etc.

Related Posts

Keep Learning More!

Sign up for our newsletter, to get the latest in SEM news and information once a month.


Submit a Comment

Your email address will not be published. Required fields are marked *