Evaluating Search Volumes for 72k Key phrases [A Study by Ahrefs]

There are two sources of key phrase search quantity information that come straight from Google: Google Search Console (GSC) and Google Key phrase Planner (GKP). We determined to match the numbers between the 2 and see how a lot they align.

GSC is extensively thought-about the “single supply of fact” for correct key phrase information as a result of it reveals you the precise variety of instances your web page appeared in Google’s search outcomes for a particular key phrase.

So in case your web page constantly ranks on the entrance web page of Google for a given search question, the variety of impressions for that question in GSC ought to precisely mirror the search quantity (in most circumstances).

As for GKP, its search quantity information is understood to be lots much less exact. Primarily as a result of:

• It teams key phrases with related that means.
• It rounds search volumes into buckets.

And but, quite a lot of web optimization professionals are completely pleased with the search volumes they get from GKP and think about them to be correct.

So we determined to run just a little experiment and research how search volumes from GKP stack up towards the “single supply of fact,” aka GSC.

Evaluating search volumes between GSC and GKP

For this experiment, we took 72,635 random key phrases within the 1K to 10K search quantity vary. Then we in contrast the variety of “impressions” in GSC with the search quantity information from GKP (for a similar month).

The very first thing we discovered is that GKP nearly at all times overestimates “true” search quantity:

However the true query is by how a lot? In any case, if it solely overestimates search volumes barely, then what’s the massive deal?

Right here’s the reply:

Most search volumes in GKP (54.28%) are overestimations, whereas just below half (45.22%) are roughly correct (i.e., deviating from GSC “impressions” by not more than 50%).

For the web optimization nerds amongst you, right here’s a extra granular view of how GSC information compares to GKP:

Fairly a discrepancy, proper?

So why don’t we dig just a little deeper and examine the precise causes of such a stark distinction in search volumes between GKP and GSC?

GKP teams key phrases with related that means (and makes errors)

In keeping with our GSC information, the key phrase “ahrefs” bought 25,436 impressions in June (within the U.S.):

But when we take a look at the information from GKP for a similar month, it reveals a search quantity of 33,100:

Which means the “GKP/GSC” ratio for this key phrase is 1.3x. Not too unhealthy, however not very correct both.

So what’s inflicting this discrepancy?

Apparently, the GKP search quantity for the key phrase “ahrefs” contains the search volumes of all its misspellings that we are able to see in our GSC:

• ahref (2,826 impressions)
• hrefs (906 impressions)
• aherfs (435 impressions)
• arefs (267 impressions)
• a hrefs (224 impressions)
• aherf (185 impressions)
• ahrfs (100 impressions)
• and so forth

We all know this as a result of GKP teams them collectively (and reveals the identical search quantity that it reveals for “ahrefs”):

However when a key phrase’s search quantity accounts for its misspellings, it’s not that massive of a deal, proper? In actual fact, it may well truly be fairly helpful.

Nicely, there are some circumstances when GKP is grouping issues that shouldn’t be grouped. And this may be deceptive for us SEOs.

For instance, the key phrase “chusky” has a search quantity of 550k within the U.S. GKP thinks that it’s a misspelling of “husky” as a result of these two key phrases are being grouped collectively:

However if you happen to take a look at the search outcomes for the key phrase “chusky,” you’ll immediately see that it’s not a misspelling. Moderately, it’s a singular canine breed, that means that it ought to have a definite search quantity of its personal.

Similar story with these 4 key phrases: “crimson room,” “crimson rose,” “crimson rock,” and “crimson robin”:

GKP treats them as one, however you don’t even should examine the search outcomes to know that these items aren’t the identical.

Just a few extra examples:

• american banks & financial institution of america
• mosquito bites & mosquito bits
• ebook a driver & drive ebook

In all honesty, these sorts of “false groupings” aren’t frequent, however they could trigger you a lot of complications must you stumble throughout certainly one of them.

What’s much more frequent (and aggravating) is that GKP doesn’t present distinct search volumes for intently associated search question variations.

For instance, the next search queries are grouped collectively in GKP with no strategy to see particular person search volumes:

• computer video games free obtain
• free computer video games obtain
• free video games obtain for computer
• obtain free video games for computer
• free computer sport downloads
• obtain computer video games free
• free video games to obtain for computer

This “grouping” challenge is then being picked up by each web optimization software that pulls its search quantity information from GKP (and almost all of them do that).

However right here at Ahrefs, we combine key phrase information from a number of sources (together with a few years of historic clickstream information). This permits us to “un-group” search queries and present the distinct search quantity for every variation:

Figuring out the distinct search quantity of every particular person search question prevents you from unintentionally overestimating a subject’s complete search site visitors potential when summing the search volumes of all key phrases in a group.

Apart from that, realizing the preferred methods of how individuals phrase their search queries will help you modify the language of your web page accordingly and give you a extra eye-catching web page title.

As you possibly can inform, this “grouping” function in GKP could be irritating for us SEOs. Nevertheless it’s not like we are able to blame GKP for hiding exact search volumes from us. GKP is a software for advertisers, not SEOs. And this grouping of comparable key phrases is definitely handy for them.

GSC reveals native impressions for search queries with native intent

In keeping with GKP, the key phrase “golf programs” has a formidable common month-to-month search quantity of 1 million within the U.S.:

However right here’s the factor. Relying on the placement of the place you’re looking from, you’ll see completely different pages rating for this key phrase:

• In case you search from Rochester, you’ll see a web page from visitrochester.com at #2.
• In case you search from Bakersfield, you’ll see a web page from visitbakersfield.com at #1.

Thus, the house owners of those two web sites will see a distinct variety of month-to-month impressions for the key phrase “golf programs” of their GSC. That’s as a result of every web site ranks nicely for this key phrase solely in a particular location.

And solely the proprietor of an enormous web site like golflink.com (which appears to have pages rating for “golf programs” in each conceivable location) would seemingly see the variety of impressions that’s near the 1 million that GKP reveals us.

These “regional” key phrases very often have the largest discrepancy (4x+) between GKP and GSC numbers, which could be seen in our graph above.

“Impressions” in GSC are generally inflated by bot site visitors

Let’s discuss these uncommon circumstances the place GSC reveals the next search quantity than GKP.

We consider that it seemingly occurs due to the bot site visitors. In keeping with John Mueller, not all the impressions from bots are filtered in GSC:

Generally it may be from bots — we don’t essentially filter all of that out in Search Console.

— 🧀 John 🧀 (@JohnMu) August 6, 2021

However what is “bot site visitors?”

Nicely, that’s any type of script or software program program that does automated searches in Google.

The “bots” that I’m positive you’re acquainted with are rank trackers that make automated searches in Google to report the place your web site ranks.

A a lot nastier instance is bots that generate faux clicks on Google adverts to place some strain on their rivals.

Anyhow, in line with our research, GSC information appears artificially inflated in solely 0.5% of circumstances. So it’s unlikely that you’ll undergo a lot from bots polluting your GSC studies.

How does Ahrefs’ search quantity information stack up?

I’m positive a few of you might be questioning how Ahrefs’ search quantity information compares.

Nicely, let’s plot “Ahrefs/GSC” ratios proper subsequent to “GKP/GSC” ratios from the earlier graph:

It seems to be like Ahrefs reveals “roughly correct” values in 60% of circumstances vs. 45% of circumstances for GKP. 

That’s largely attributable to our potential to “un-group” clusters of comparable key phrases and report distinct search volumes for every of them.

So if you happen to had been questioning why the search volumes in Ahrefs aren’t the identical as these in GKP, now you recognize that’s by design.

Last ideas

I hope you loved this analysis research and that it gave you a greater understanding of how the numbers in GSC and GKP differ and, extra importantly, why they differ.

Have questions? Ping me on Twitter.

And by the way in which, big thanks to our superior information science staff for carving out a while of their busy schedules to assist me with this analysis.