25th June 2020
28th May 2020
26th May 2020
6th May 2020
6th May 2020
15th March 2020
18th February 2020
24th January 2020
Culture - 14th June 2019
Culture - 29th March 2019
A creative agency with deep digital DNA
MAKING SENSE OF GOOGLE’S LATEST CHANGES TO KEYWORD PLANNER
Has Google’s latest Keyword Planner update left you and your Search team scratching your heads in bewilderment? Here, Huskies Lead SEO Analyst Conor Mulcahy offers advice on how to make sense of this change, and asks why a company with such talent and influence chooses chaotic solutions to relatively simple problems…
In the summer of 2016, Google rolled out two changes to their Keyword Planner tool – the introduction of keyword volume ranges and combined volumes for close variants. Unfortunately these changes are not helpful for digital marketing agencies or clients alike. These changes make the keyword research stage a much more laborious task for the digital marketers, SEO & PPC experts who are aware of the change and for those who remain unaware, they will likely set inaccurate forecasts based on poor data. With that being said, it’s important to be clear on these changes and understand how we can adapt to our new surroundings.
Since it’s inception in 2008, Keyword Planner was intended for both media planning and research. Yet it would seem now that the tool is becoming less suitable for research purposes and more a gateway to purchasing ads directly from a registered email address. This is a nuisance for agencies, as not every agent will be running the campaigns from their own individual accounts.
For a number of years, people like ourselves (SEO’s & digital marketers) who don’t run Adwords campaigns through their own email account will be familiar with the following message in their Adwords interface:
The observed red bar essentially advised users to do things:
– Add credit card details to the Adwords account
– Spend money by setting up and running campaigns
AdWords users who are used to seeing this red bar, with lower monthly spends or without live campaigns, are now likely to see lower data or data ranges in the planner. The ranges are as follows, and as you can see they strongly affect the accuracy of the search query volumes:
At the beginning of August, I was greeted with the below view when logging into the Keyword Planner. Now, it must be said that not everyone in Huskies had been shown these ranges, but many had.
As a search marketer, I use this tool every day. I was forced to find a solution pretty fast.
If you have an existing email address that doesn’t run Adwords campaigns, it looks likely that this account will not be suitable for your research. Even when credit card details and campaigns were added to my existing account, the ranges remained
The solution I discovered was to register a new email address in Adwords, inputting credit card details and then setting up a campaign – which was then paused.
First spotted in June, Google began merging keyword volumes for close variant keywords. This led to search marketers across the globe overestimating search queries, as Google was attributing the volumes to multiple keywords.
Google later acknowledged that volumes are combined for “close variants”, yet how a close variant is distinguished remains up for debate. Though thesempost.com believes them to be:
– Plurals with non-plurals for any word in the keyword phrase
– Acronyms with longhand version
– Stemming variants: -er, -ing, -ized, -ed etc keywords (ie. designer, designing, designed)
– Words that can be spelled with or without space (ie. car park and carpark)
– Words with and without punctuation (ie. kid toys and kid’s toys)
Most of the keyword tools that digital marketers use on a daily basis extract data from Google’s Keyword Planner one way or another. With this being the case, the data (for now) will likely be inaccurate.
My advice is to conduct your keyword research as you normally would, then add the additional step of re-running the numbers through the keyword planner, as shown below.
Let’s be clear about these volumes before we run through the solution. We are not comparing apples with apples here, nor are we comparing apples with oranges. The comparison is somewhere in between. I’ve amended Google’s official definitions to interpret the difference, as these official definitions remain outdated.
Average monthly searches: The average number of times people have searched for a combination of similar keywords based on the location and search network targeting selected. By default, we average the number of searches for the term over a 12-month period.
Estimated Impressions: The number of impressions divided by the total number of searches from the location and network for a given month. Estimated impressions take the bid, budget, and historical ad quality into account, but search volume statistics don’t. Impressions are also triggered when a user clicks beyond page one of the SERP’s.
There’s no new instructions to heed for Step 1. Simply add your keywords, assign the country and run the query.
It’s at this stage that we see the inaccuracies caused by the grouping of close variants. From henceforth, we must advance our keyword research by forecasting the numbers as you would for a PPC campaign.
To do this, simply add the keywords to your plan by pressing “add all” (or by adding individually), followed by pressing “review plan” button.
Enter a relatively high bid to set the Keyword Planner budget to get maximum impressions.
Set the custom time period to the duration of your campaign or to a full year if you wish to gather average monthly keyword volumes, then choose “monthly” from the drop down menu.
The limitation here is that we cannot foresee seasonality trends, which will impact on traffic forecasting.
Select “exact” from the match type.
To view the keywords, click on the “keyword” tab at the top of the page. This view will display the keyword volumes for each of the keywords. Considering I added a high bid in the previous step, the average positions are all very high – position #1 in fact.
Sort the keywords by impressions to view the keyword volumes (impressions).
Now that we have the data, we’ll need to download it.
– Deselect “Historical statistics”, as this uses keyword volumes and not impressions.
– Select “Traffic forecasts”
– Download Excel CSV
Once downloaded, the estimated impressions can be opened in Microsoft Excel or Google docs and these provide a more accurate representation of actual query volumes.
This data is what we’ve been looking for, though we have gone through some unnecessary obstacles to get it.
It will become more of a problem when we try to scale the ideation of large sets of keywords. Researchers will be inadvertently excluding valuable keywords and including less important keywords due to the falsely represented volumes, which have been double attributed by closely matched variants.
We already have the data, so there’s no need to run this step for campaigns. But I thought it would be a worthy pursuit to compare the volumes of monthly searches using the traditional method versus estimated impressions using our new method (fundamentally a PPC method) – just to see how much of an impact this could have on the projections of a less-informed digital marketer.
We can see in the graph below that there is little correlation between the two sets of estimations. For instance, the average monthly searches of “car leasing” & “car lease” is approximately double the estimated impression, yet the projections are roughly the same for “lease a car”. The last key term on the graph “lease car” overestimates the number of monthly Searches by 350%.
With a simple comparison such as this, we can see how the estimates compare and it’s worrying to think how many people will be deceived by this change. For a company so focused on metrics and user behaviour, search marketers could be vindicated for their suggestions of hypocrisy.
By combining the totals of these 5 keywords, it becomes clear that the tried and (formerly) trusted method of keyword research is over estimating query volume in this instance by as much as 85%.
For those of us who have worked in this industry for many years, changes such as these come as no surprise. Google has a history of obfuscating data and making significant changes with little to no warning at all.
The data is still available to some extent, even if we are comparing apples with pears. Here’s what Google said:
“This change was made so that we can consistently give advertisers the data they need to optimize their accounts, while preventing “bots” and other services from abusing the intended use of Keyword Planner. The search volume estimates, though displayed differently in some cases, still provide an accurate and helpful view of how many clicks and impressions keywords may receive.” – Google Advertiser Community
In reality though, their response is hardly believable, as there are much better ways to prevent bots from abusing the tool. A company as smart as Google shouldn’t be limited to such archaic methods as serving inaccurate data. The search volume metric has, in effect, been made redundant. But yet it remains the most visible metric to advertisers within Keyword Planner.
If they truly wished to consistently give advertisers the data needed to optimise accounts, wouldn’t it be prudent to replace the average monthly searches metric with estimated impressions, similar to how it used to be? In doing so, campaign projections would be at least somewhat realistic and search marketers less exposed to their inaccuracies.
There’s certainly plenty of scope for how Google could improve the experience for their advertisers, but for now, it’s up to us to evolve and adapt. Again.