How Accurate is Google Trends?

When you are worried about something: what do you do?

When you like something nobody knows: where do you look for information?

When you feel pain: where do you look for information about it?

 

The answer is: We search on Google about it.

Google knows you better than your wife, your mother or your brother.

Think about it:

  • If you look for a hangover remedy, it immediately knows that yesterday you had a remarkable party.
  • If you look for certain cooking recipes, you are giving it information about your gastronomic preferences.
  • If you look for your favorite sports team results, Google knows that you’re a supporter.

 

Moreover, we search on Internet without any filter at all. We don’t think twice.

Everything you look on Internet leaves valuable information about what are you interested about, what is actually worrying you… and even your deepest and darkest desires.

If you now take all that information and multiply it by the number of people living in your country, you have now a priceless Market database.

 

If now, you take all the information about all the countries in the world and put it together, you have Google Trends.

Google Trends offers unlimited free data about what people look for on Internet worldwide.

And this information is waiting for you to interpret and use it.

 

Here, we will provide you with:

  • Advice on  how to search this data.
  • Tips for obtaining the best results possible.
  • Practical examples with real data and conclusions.

 

Let’s begin:

How do I use Google Trends?

The first step is to answer the questions:

  • Which data should I collect?
  • What should you look for?

 

As we mentioned before, Google Trends offers unlimited data.

Choosing the right one is the first step to obtain valuable results.

How to obtain good data on Google Trends

You should consider 4 things when looking for data in Google Trends:

  1. Be sure about what you are looking for.
  2. Don’t look for results that confirm what you think.
  3. Look for tendencies rather than absolute values.
  4. Start with long-term data and then move towards short term.

 

Lets explain them in more detail:

1. Be sure about what you are looking for

Looking for information with just a vague idea about what you’re looking for is one of the biggest mistakes you may commit.

 

You must be open to find things you didn’t expected, of course, but “surfing” the data without any destination rarely ends up in good results.

2. Don't look for results that confirm what you think

Flat vs Round earth Google Trends legend

Check how there are more Google searches about “flat earth” than “round earth”.

 

This is known as “confirmation bias”.

If you want to find data that confirms what you think… you’ll always be able to do it.

Even today, you can find people who think that the earth is flat.

 

So, look for patterns, correlations… but in an open way.

3. Look for tendencies rather than absolute values

Google Trends Search Console and Analytics legend

Check how the global interest on Google Analytics has been declining while, on the other hand, people is becoming more interested on Google Trends and Search Console.

 

If a certain Query presents an upward or a downward trend during years, you’ll be able to guess whether people is being more or less interested about it.

 

By using tendencies, you’ll be able to get ahead of the market and obtain more solid conclusions.

4. Start with long-term data and then move towards short term

Solid Data is consistent during the time.

  • You should never look for a 1-month correlation.

 

As soon as you confirm a long-term correlation, you can look for more accurate data reducing the time period.

 

We’ll give you two examples: One of a strong correlation and one of a weak correlation.

Strong Correlation vs Weak Correlation - Google Trends example

Solid correlation – Google Trends

 

Check the correlation between “Christmas decorations” and “Presents“:

 

It is a solid long-term correlation that, of course everybody could expect.

 

However, what else can we obtain from this analysis?

If we reduce the period from 5 years to 1 year, we have more precise data. We observe that:

 

  • The peak of interest on “Christmas decoration” happens on 2 – 8 December.
  • The peak of interest on “Presents” happens on 9 – 15 December.

 

This is relatively obvious but also very interesting:

We confirmed that we tend to worry about the Christmas decorations before we do about the Presents.

 

You could use it:

  • For creating a marketing campaign that said, for example: “We also tend to be more worried about our Christmas decoration than on the presents themselves, we understand you…. bla, bla, bla“.
  • Placing Toys promotions close to Christmas decorations.
  • etc, etc.

 

Weak correlation – Google Trends

 

Check the correlation between “party” and “ice cream” on a 30 days period of July – August:

 

As you can appreciate, it seems that, as soon as the interest about “ice creams” rises, so does the interest about “party”.

  • Is there any correlation between both concepts?
  • Maybe when people eat ice creams then feel the urge of partying hard?

If you check those “peak” days, they are weekend. Nothing else.

 

Moreover, if you take a whole year period:

 

There is no correlation at all.

This last example shows one very important thing that you should never forget:

  • Good results make sense.

How to interpret Google Trends data?

Once you have obtained interesting and unexpected correlations, you should use your common sense.

Ask yourself:

  • Does this correlation make any sense?
  • Is there any good explanation?

Unless you can find a logical explanation for the correlation you’ve found, we wouldn’t encourage you to settle with it.

You have to be careful.

With all the information available on Google Trends it is easy to end up looking for a “miraculous” correlation between “Chicago weather” and “tuna price on Tokio”.

 

Don’t get over-excited and use the results obtained with common sense.

 

Sometimes, only sometimes, you may find a “hidden” unexpected good correlation few people noticed about.

And even those correlations are connected somehow.

  • You only have to think harder.

How to use Google for Market Research

Now it is time for us to give you some practical examples about useful data you could find using Google Trends.

 

In addition to these examples we’re about to explain, in our “SEO section” we have talked about what we use Google Trends for:

 

Lets see some interesting examples:

Samsung or Xiaomi? - How to use Google Trends

Imagine that you had a Blog that analyzed different mobile phones.

Historically, you had been focusing on Samsung mobile phones successfully.

  • New models.
  • Last features.
  • Prices, etc.

 

However, you started to realize that your users are no longer interested in your Blog.

  • Your website traffic decrease little by little each month.

 

You then decide to check Google Trends in case you find some answers.

After some research, you find out this (Worldwide search):

 

While global interest on Samsung has been declining last 5 years, the overall interest on Xiaomi increased drastically, even surpassing it.

 

For you, the Solution is simple: You have to talk more about Xiaomi mobile phones.

 

This doesn’t mean that you should forget about Samsung mobile phones.

  • In fact people could become interested again in next months.
  • Google Trends just gives you Trends, not the future.

 

What this analysis tells you is that you forgot talking about other emerging brands.

Burger or Taco restaurant? - How to use Google Trends

 

Now, you are planning to start a restaurant but you’re not sure if you should start a Taco or a Burger restaurant.

  • Since it implies a physical location, lets assume you live on Chicago.

 

You are “pretty sure” that US population is more interested in Burgers rather than Tacos. In the end, what is more American than a Burger?

 

Since you don’t want to risk your money, you decide to check Google Trends.

  • You decide to use a 2004 – present period of time.

 

What you find out is astonishing (US results):

 

Americans (US) are more interested in Tacos rather than in Burgers.

 

But what about Illinois?

 

People from Illinois are even more interested in Tacos than an average American (US).

 

However, you need more proof. We are talking about your personal savings.

You then decide to compare the interest of Illinois population in 2 famous restaurants that elaborate Burgers and Tacos:

  • Burger King and Taco Bell.

 

We are aware that Taco Bell makes everything but Tacos.

  • However, it can be a good way of measuring people’s interest in those particular dishes.

 

You find out this:

 

You have no doubt: your restaurant will be serving Tacos rather than Burgers.

Which city should you visit? - How to use Google Trends

 

This example is a bit funny, but it perfectly shows how you should handle certain Google Trends results.

 

Now you are planning a trip to a European city.

You have narrowed your list down to 4 cities you’d love to visit: Rome, Berlin, Barcelona and Lisbon.

 

But, there is a problem: you hate over-crowded touristic destinations.

 

Then, you decide to use Google Trends in order to find out how many people is looking for visit each city.

  • Since you are planning your travel for next month, you don’t worry about past results but the current ones.

 

Suddenly you realize about something: it is not the same having lot of people on a big city than on a small city.

  • 50.000 tourists would be much more annoying on Venice than on London, for sure, since Venice is much smaller.

 

What do you do?

You decide to normalize that Google Trends results by the different cities’ populations.

  • You just take into account the city-centre population since nobody visits the suburbs.

 

Then, you obtain a city touristic-crowding factor:

  • Rome: 2.87 Million people. Interest (78)/ Population (M) = 27.1
  • Barcelona: 1.6 Million people. Interest (71)/ Population (M) = 44.3
  • Berlin: 3.75 Million people. Interest (59)/ Population (M) = 15.7
  • Lisbon: 0.545 Million people. Interest (26)/Population (M) = 48

 

Check out how interesting these results are:

  • At first glance, Lisbon seemed to be your city.
    • However, since it is a relatively small city, compared to the other 3, its Interest/Population factor makes this city the “worst”: 48.

 

  • On the other hand, Berlin presents a very low factor15,7due to its big size and the moderate Google Trend interest in it.

 

Conclusion: You should pack your things and buy a ticket to Berlin.

Summarizing

Google knows us better that anybody.

Our Internet searches represent one of the most sincere, valuable and limitless sources of information we’ve ever had.

 

Google Trends allows you to access this information in a free open way.

If you want to get ahead your competitors, you should start caring about all this data.

 

In order to obtain valuable data successfully from Google Trends, you should:

  1. Be sure about what you are looking for.
  2. Don’t look for results that confirm what you think.
  3. Look for tendencies rather than absolute values.
  4. Start with long-term data and then move towards short term.

 

You should never forget one important detail when interpreting that data:

  • Use your common sense.

 

No matter how impressive the correlation you’ve just find is:

  • It should make sense.

© 2022 - Consuunt.

CONTACT US

We're not around right now. But you can send us an email and we'll get back to you, asap.

Sending

Log in with your credentials

Forgot your details?