When exactly is something not exact? When it’s exact match... .


In the distant days of long ago, there used to be a time when ‘exact’ meant exactly that. To become an exact match, a keyword search had to be exactly the same, and ‘close variants’ were introduced to include the likes of variants, plurals, typos etc. Advertisers were able to opt out of that until several years ago, but Google has since removed the ability to do so.

We saw another change from Google just last week – from April, the close variants will now ignore function words such as ‘in’ and ‘the’ when they do not change the meaning of the keyword, and will also ignore any reordering of the words… all the while, still classing them as ‘exact matches’.

So what exactly is the change?

Function words

As Google explains, function words are binding words such as prepositions (in, to), conjunctions (for, but), articles (a, the) and other words that often don’t impact the intent behind a query, i.e they do not change the meaning of the keywords. With this change, as they do not affect the meaning they can be ignored.

However…there will be cases when they will not be ignored, the example given that ‘flights to new york’ is not the same query as ‘flights from new york’. Below are some examples of from Google of matching search queries with function words which have been added, removed, or changed in the query.


Word order/Re-ordering of words

In English, word ordering doesn’t always make a difference. This has led to users often not using a natural word order when searching as the intent is the same.

Take a keyword like [buy electric guitar]. The meaning doesn’t change with [electric guitar buy] or [guitar electric buy]. Exact match will use logic to match with queries that are reordered variations of your keyword.

One thing that won’t happen is that note word reordering won’t add any words to your keywords, and keywords also won’t be reordered to match with a query when it changes the original meaning of those keywords.  As a further example, Google state ‘For example, the keyword [SFO to JFK] shouldn’t match to the query “JFK to SFO” because the destination is different.’


Why is it happening and what does it mean?

This latest version of the definition of ‘exact match’ means that Google is placing increasing trust in its machine learning and the belief that it’s ‘now at the point where advertisers can let the algorithms take over and focus on other things.’ It would appear that the philosophy is to spread a wider net, and then filter out what you don’t want, rather than build a net that might not be big enough to catch everything you want.

These changes do not apply to phrase match keywords. And AdWords is still designed to prioritise matching identical keywords to identical search queries.

It also means that granular lists of exact matches are no longer required, you won’t for example be seeing differences between how many users searched for and clicked on ‘running shoes’ compared to ‘shoes for running’. Whilst this may make our jobs simpler in a way, it does take away control from us, and it does this by bringing less control over bids and reduces relevancy, and reduced relevancy will mean wasted spend and less Return On Investment.

Take the example above about ‘running shoes’ v ‘shoes for running’. When analysing the ad spend it may have been proven that ‘running shoes’ only costs for example 50p a click, whereas ‘shoes for running’ costs £1 a click. Now the control for which to bid for will be gone.

Another query could be ‘running shoes’ v ‘shoes running’. It’s a very similar query but there could actually be a difference in what the searcher was looking for and so the click could be wasted spend. The reason? If the searcher was actually looking for shoes which are running then the £1 click is a wasted click (or, as a few people in the industry are beginning to pick up on, it’s more money to Google..)

So what do we feel about this, what will the impact be?

Advantage Google... Inappropriate variations of search terms are often only discovered after an advertiser has paid for wasted clicks from variations showing up in a search query report, which as mentioned above brings ROI and budget spend implications.

Precision control is being handed over to machines via automation, and so advertisers/execs will have to be increasingly diligent about using and mining search query reports, and thinking ahead about unintended consequences when word order matters. It would seem the focus is now on what queries advertisers don’t want their ads showing up on rather than what they do want.

Google estimates that “advertisers may see up to 3% more exact match clicks whilst maintaining click through and conversion rates.” (Source - https://adwords.googleblog.com/2017/03/close-variants-now-connects-more-people.html)

Ways to prepare for and manage the changes

There are a few things which can be done to prepare for the coming changes.

  1. Review existing exact match queries and determine if the loss of function words or a reordering of the words changes the meaning. Add those variations as negatives in your campaigns.
  2. Review close variants in your Search Query Reports to see if other variations are currently being triggered that might be affected by these changes. Add those as negatives.
  3. Starting in April, step up your mining of Search Query Reports, particularly for close variants.

In other words….be prepared to expand your negative lists even further!!!



29 Mar 2017 - 6 minute read
share this

stay in the know, stay ahead.

Get the latest from the agency, including news, events and expert content.
explore services in the article
find out what we can do for you
read some of our case studies