Google Expands Broad & Phrase Match Types for All Advertisers
Darko•August 12, 2019
Google Ads are Changing Phrase Match and Broad Match Modifier
Through a blog post, Google Ads have announced that phrase match and broad match modifier keywords will be changing in the coming weeks. These will start to include search queries that share the meaning of keywords being targeted by advertisers. This is an extension of what Google has previously rolled out for exact match keywords, so in a way this will not come as a surprise. However, this will likely be frustrating for much of the more advanced PPC community and may pose a risk to accounts that are either poorly structured or not managed closely enough.
A Brief History of Google Ads’ Exact Match Changes
Let’s understand a little bit about where this came from. Back in 2014, “exact match” keywords were exactly that. If you targeted [keyword], even the plural [keywords] would not show an ad if you specifically selected an option in your campaigns that prevented you matching against “close variants”, which included plurals and misspellings. In essence, back then, exact meant exact.
Since those days, Google have made several revisions to exact match:
In mid-2014, Google required that all advertisers use “close variants” which would capture misspellings, plurals and variations of the same word. Keywords such as “shelf” and “shelves”, “clothes” and “clothing” and so forth could now match against the other.
In 2017, Google then extended close variants to ignore function words where appropriate and to also accept re-ordering of words where it detects the same meaning.
Some Issues PPC Marketers Experienced in the Past (and Now!)
On the surface, these changes sounded like a good way for consumers to reach more searches and more customers without having to be extremely exhaustive and target thousands of slightly different queries. However, this ended up producing more work and management time amongst the professional community who more carefully manage their accounts. Some examples include:
Plenty of plurals such as “glass” and “glasses” can have different meanings when used in different contexts
Brand marketers in particular found reordering of keywords and ignoring function words could have a detrimental effect: “the piccadilly hotel” may be your fenced off brand territory, so starting to match against “hotel by piccadilly” would bring all sorts of different user intent that required extensive negative management
There are many cases where Google’s machine learning will often not get things right when it comes to similar intent: examples include cable cars that include “360” in their name being matched against products like 360 degree cameras. Study related keywords such as “bible study” and “biblical studies” can also often mean different things
I should caveat these examples by saying Google’s AI tries constantly to match user intent, and therefore the above examples should “fix themselves” over time. However, given that so many new searches are made on Google every day, it’s impossible for this to be 100% correct. This will inevitably lead to advertiser wastage (and additional revenue for Google).
The logic behind these past changes is that due to different ways we search differently (voice search being one prime example), a decent percentage of all search terms are brand new. Google says this is as high as 15% on a daily basis, and we can’t expect marketers to anticipate all of these new searches for everybody. I agree that the sentiment here makes sense, but what Google doesn’t say is that these changes could also lead to advertisers spending a lot of extra money while collecting that new keyword data.
Okay, I Get the History. What’s Changing Now in Google Ads?
An ongoing joke within the PPC community following these historical changes was that because the above changes only affected exact match, “phrase match is the new exact match”. Essentially, if advertisers wanted to avoid needing to deal with a term matching against a large number of things that they didn’t originally intend, phrase match was arguably more precise in many circumstances. However, the joke is now about to die as Google rolls out all of the above changes to phrase match and broad match modifier terms as well.
The same meaning queries with exact match will be rolled out to both phrase match and broad match modifier terms over the next few weeks. An example of this is “lawn mowing services” where on these match types, you mind find yourself now matching against similar queries such as grass cutting services. Below are examples for both broad match modifier and phrase match:
What Does This Mean for Advertisers Using Phrase Match and Broad Match Modifier?
For highly structured PPC accounts, this change will likely come as a headache. Being able to specifically pinpoint where a query will go in the account is important, and sometimes single-keyword ad groups or even campaigns are the way forward for high-spend advertisers wanting a high degree of control. Therefore, having more keywords suddenly start to match against other things will cause advertisers a headache in terms of finding and rooting out negatives.
When accounts are less structured or are younger, the effects perhaps will not be felt as widely. However, it will be important for everyone to be constantly checking their spends and spotting where their increases in queries goes up more than they would like.
What Should I Watch Out For?
Google estimates that most advertisers will see a 3-4% increase in clicks in broad match modifier and phrase match campaigns, with 85% of these new queries coming from newly expanded search terms. If you see your campaigns spike in spend considerably more than this, consider diving in and pruning back sooner rather than later.
Just like after all fundamental changes to the way Google Ads are run, make sure you are monitoring performance and perhaps put a few extra reminders to check search query reports over the coming weeks. It is likely that larger, high degree of control PPC accounts will be affected more by this change, and may have to update their structures accordingly.