Once upon a time — aka 2013 — if you added a keyword to your Google Ads account as “exact match,” it meant that your ad would only trigger if someone typed in your keyword exactly. Makes sense right? “Exact match” = match search term exactly. However, one day in 2014, SEM managers across the world woke up to a wrench in that equation. Since then, search query reports have never been the same.
The Life Story of Exact Match
2014 – Similar Variants – In the first update to exact match, Google introduced similar variants. This generally meant that misspellings and plurals would match to your exact match keywords. While frustrating at times, this update was often helpful, as you no longer had to create a keyword for each misspelling of your brand name or product offerings.
2017 – Different Word Orders – For the next three years, you could continue to trust that your exact match keywords would be matching to search queries that either matched exactly or were very close to your keyword. That is, until 2017, when exact match expanded to include different word orders.
While generally not a huge issue for search advertisers, this change did result in the occasional issue. For example, if your keyword was New York Life Insurance, there was clearly other intent in someone searching for “life insurance new york”. However, these queries were getting matched to the exact match keywords.
2018 – Intent – Google recently announced that exact match keywords would begin matching to search queries with similar intent. While Google has become exponentially more sophisticated in their ability to leverage machine learning and recognize patterns in intent, we’re seeing some clear flaws in this update.
In one example, we’re seeing that the exact match keyword “workers compensation” is matching to the search query “work compensation.” You can see how this is problematic and can quickly turn into spending funds on irrelevant terms. In the case above, if you’re selling workers compensation, you do not want to be spending money on clicks where a user is looking to do research on salaries.
Original Article Here