All Collections
Rank Tracker
Accuracy-ensuring features
Google Ads search data calibration systems
Google Ads search data calibration systems

Avoid being misled by Google Ads search volume data when using it for SEO purposes (rather than Paid Search directly)

Updated over a week ago

SEOmonitor makes sure the search data pulled from Google Ads is processed correctly within the Rank Tracker and would not mislead users.

How does this work?

Google Ads search volume data is reliable, and it’s the only trustworthy source available. However, it can easily become misleading when used outside Google Ads’ platform. Our system detects the 2 possible errors and processes them accordingly.

Close variations

Google Ads aggregates the search volume of all misspellings and close variants under one main keyword. For instance, it shows the sum of the search volumes for “books”, “book” and all of their misspellings (boks, blocks, etc.) under the main keyword "books". It shows the same summed-up search volume for "book" and "boks" as well. When you find close variants like "email marketing", "emailing marketing" and "email marketer" showing an estimated monthly search volume of 14.800, you should know that the number represents the combined search volume of all three.

This means that, if a user tracks both “book” and “books”, they would get the combined search volume for both keywords, in turn, doubling the total search volume for those keywords. This distorts keyword group-level metrics, such as a keyword list’s total search volume and a website’s Visibility on it.

Our systems detect this by analyzing:

  • the 12-month average search volume,

  • monthly search volume trends and

  • top-ranking landing pages for all keywords in the campaign.

Based on that data, it aggregates the keywords, just like Google Ads does it.

By default, only the main keywords are listed in the Rank Tracker, with an indication of the number of close variants ("+1", "+2", etc.). You can see the close variants by clicking on the number icon.

We track rankings on all keywords, no matter if they are aggregated/clustered or not. The aggregates only share their search volume.

You can temporarily unnest/ungroup them to view the close variants in the list of keywords. To do this, select "Ungroup" from the "Close Variations" menu:

On ungrouping, the main keyword and variations are marked as MAIN and VARIATION, respectively. This way, you can analyze all keywords, both the main keywords and their variations, each with their own sets of data (SERP data, search data, rankings).

However, when using this ungrouped view, you should always remember that the search volume represents the sum for all the variants and not a value for each.

Sudden 0 search volumes

Google Ads does not let their users advertise on keywords that are considered dangerous, like “cannabis” or “bitcoin”, or can block other keywords for reasons similar to this one, like blocking ads on baby-related products. When they don’t let their users bid on them, they also don’t show the search volume data.

These lists of products and/or services is dynamic. Some go in, and some go out.

Now, with the search volume being an average of the past 12 months, it can’t get to 0 in less than 12 months. So when that suddenly happens, we don’t overwrite the last non-zero value we got and we keep that until it gets a new non-zero value.

However, it comes with a "warning" instead of the Year-over-Year trend, and the search seasonality is removed as well. The tooltip for the warning would explain the situation and show when was the last time when the value was updated. We already know it’s not “0”, and overwriting the search volume with “0” would erroneously affect the Visibility metric.

As future development, we will enable our users to manually overwrite the search volumes (only in this situation) with estimates from other tools, like Google Search Console, or even Clickstream data. Even though they are not as accurate as the Ads data, in some cases they might be better than old or no data at all.

Did this answer your question?