Google Updates Close Variant Matching

Google had a little Saint Patrick’s Day surprise for marketers invested in SEO, AdWords, and the like: they announced on the Inside AdWords blog that they will soon be making some changes to how the search engine uses search terms, expanding current practice to include reordering and rewording of the search terms, in order to deliver more exact results for searchers.

This update will also negate the impact of the so-called function words—prepositions (in, to), conjunctions (for, but), articles (a, the), etc.—so long as that doesn’t alter the meaning of the initial search. In this post, we walk through the why, when, and how of the change, make some recommendations regarding next steps for marketers, and weigh in on whether or not this is panic-worthy (spoiler alert—it isn’t!).


Image from Inside AdWords blog

What Does the Actual Change Look Like and When Will It Happen?

For many years, close variant matching was the default (optional) setting for the search engine. In the fall of 2014, it became mandatory for advertisers. This change simply takes the existing allowed close variants—common misspellings, singular and plural forms, acronyms, stemmings (-ing), abbreviations, and accented spellings—and adds to them word reordering and rewording (filtering out the function words as mentioned above).

Going forward, Exact Match will be somewhat like Broad Match Modifier (BMM), in that both match types can now trigger ads for queries where the search terms are jumbled in any conceivable order. Exact Match differs from BMM, however, in the sense that the query could be [sandals block heeled] and the exact keyword triggered could be [block heeled sandals], which are pretty similar. In BMM, however, the query could be [size 7 blue suede block heeled sandals for women] and the BMM keyword triggered could be +block +heeled +sandals.

It’s important to note that this change is mandatory, and advertisers cannot opt-out. It will be rolled out beginning in mid- to late-April of 2017, with a small percentage of traffic and scale from there until all traffic has been subjected to the change.

Should Search Marketers Panic?

In a word: No. Panicking about function words is unnecessary, although word reordering may seriously change the semantics game, Google has indicated that terms that have the same meaning, but are in a different order, won’t match out to word re-orders that insinuate a different meaning. It remains to be seen, however, if machine learning will really be able to pick up on subtle human nuances of language within the search terms themselves. Marketers shouldn’t panic, however they should remain alert and prepared.

Savvy marketers won’t be thrown by this change, as they are already revising negative lists and ensuring that they are built up for exact match campaigns and ad groups anyway. The biggest concern for some is that we are in a position where we have to trust in the machine-based learning of Google going forward.

This will most likely not drastically change campaign performance. CPCs may increase slightly during the transition and we can expect impressions to increase, but CTRs should still remain healthy, as matched terms won’t be drastically different.

Why Did Google Do This?

Google always aims to make the user experience as hassle-free and direct as possible. That being said, some reason they may have done this is to alleviate the pressure of building out extensive keyword lists to cover every search term imaginable. This allows marketers to reach those queries they didn’t think to build out to, connecting the searcher to their desired results.

Another reason, building of the previous point, might be to increase traffic to your campaign. As previously stated, results will show up for queries that marketer’s didn’t think to build out. This helps the advertiser reach more potential customers and helps Google benefit from the ad cost increase accrued.

Another small factor that might have lead to this decision is the issue that as AdWords continues to mature and add functionality through campaign types, ad extensions, various bid modifiers, etc., account limits become crowded for the larger advertisers. We’ve seen accounts run into issues where we have so many keywords, negatives, extensions etc. that we hit limits so we can no longer expand. By not having to build out all of these keywords, that problem is mitigated. 

How to Respond – AKA Actionable Steps

The first actionable step that most savvy marketers will take is to re-evaluate their keywords to make sure that they function during re-ordering without a change of intention, ensuring that their negative lists are updated so there should be little to no issue. Additionally, you will want to:

  • Monitor Your Search Query Reports: As with any Google update, the details are very vague on what this match type change will do to keywords. You may either see a few “close variant” queries come through on your brand terms, or you could start seeing an abundance of random queries matching your keyword—though this seems unlikely.
  • Refine Your Keyword Coverage: Based on findings within the search query reports, you may want to look at reevaluating your current keyword coverage and scrub the keywords that overlap with query mapping. You will definitely want to avoid having multiple keywords active that can pick up the same query in order to prevent any CPC increases.
  • Update Your Negative Keyword Lists: You will want to take steps to ensure that totally irrelevant queries that are being picked up by your exact match keywords will be negated across all your applicable campaigns. While it is counterproductive to go crazy with adding in too many negatives—potential loss of traffic—you will want to continue making sure that you are targeting qualified queries to avoid any potential CPC increases.

The Takeaway

It’s a big change, and it will definitely affect how you continue to do business. But, if you take the correct actionable steps, you will be fine. Eventually, as the machine learning catches up to the nuances in language that users can push through a search field, we may begin to see this change enhancing our traffic and search results over time.

*Special thanks to our Paid Search team for compiling this breakdown, especially Kayla Frawley and Hana Lee for their helpful tips and insight.

Blog Categories
Paid SearchE-CommerceIntegrated MediaCompany UpdatesPaid SocialShopping & FeedAmazon and MarketplacesSearch Engine OptimizationConversion Rate OptimizationPerformance DisplayInterviewLead GenerationUncategorizedPressIndustry NewsMobileFunConferenceEvents
Blog Archives
2017 2016 2015 2014 2013 2012 2011