On February 4, 2021, Google announced that it would be deprecating broad match modifier keywords and rolling their functionality into phrase match targeting. It’s a huge announcement, so let’s explore what it means, the implications for performance, and the timelines everyone should know.
Broad match modifier keywords, or BMM for short, were something of an unintended phenomenon in paid search when they launched in 2010. In their purest and intended form, they were designed to anchor a specific term within a broad keyword by adding a ‘+’ to the front of the word to tighten the matching options. It allowed the keyword ‘+miami vacation rentals’ to match to queries related to vacation rentals, like queries about hotels, but kept the term ‘miami’ from matching to related terms like Ft. Lauderdale or Florida, which would happen with standard broad match.
BMM as Google originally intended
Broad match often required generic ads because the query would rarely match up 1:1 with the copy. BMM was Google’s olive branch to advertisers who (rightly) claimed broad match was wasteful and unreliable. That’s the way BMM was sold to the industry, but advertisers quickly saw an unintended and better use of the ‘+’. Performance marketers realised that appending a ‘+’ to every term in a keyword created a more powerful and useful keyword match type. It removed the need for a query to match word order, as in phrase match, and still limited Google’s ability to match a query only to those that held every desired term. Thus, BMM was born and widely adopted as an industry standard.
BMM as advertisers have enjoyed since 2010
The result gave advertisers more flexibility matching to relevant queries but still allowed for control over what ad copy served against specific queries. Consider the example above. An ad with the copy “Get 10% off our best rates on Miami vacation rentals when you book for 2 or more weeks” has a strong value proposition and call-to-action, but might be wasted for a search on Miami hotels. The new BMM kept this from happening as long as an advertiser had set up their campaign correctly. It also left phrase match all but useless outside of the most nuanced of scenarios.
Google is changing the way phrase match connects queries to keywords. In the past, a phrase match keyword like “new restaurant” would need to appear in a query as written, although additional words could appear before or after the phrase. A query like ‘new restaurants near me in Brooklyn’ or ‘Chicago new restaurants’ would trigger the aforementioned phrase match term. As of mid-February, that same keyword (“new restaurant”) will expand to queries like ‘new Thai restaurants’, ‘restaurants in New York’, and so on. For those advertisers who have continued to use phrase match, this will expand the impressions those keywords see and likely decrease conversion rates on higher volume.
As of July, no new BMM terms will be permitted. Technically, BMM has never been a keyword match type within the engines, so nothing will look different in editors or performance grids, but advertisers will no longer be able to anchor terms within new broad match keywords.
This change will bring further hand-wringing within the paid search industry over the loss of control, the forced march towards automation, the loss of query data, and the “keywordless future” we have all foreseen. I will spare you those takes here. While all of that may be true (it is), this is not the most egregious example of any of those issues. Nevertheless, here are some suggestions on what to do—and what not to do—in light of this new reality
This almost goes without saying, but if you’re an advertiser using both BMM and phrase match keywords in your account, it is decision time, and one of them has to go. Why? Because both existing phrase match terms as well as existing BMM terms will now be eligible to serve on the same queries.
In all likelihood, this will result in a loss of impressions for your BMM terms and a spike in phrase. It wouldn’t be so bad if the same query always mapped to the same keyword going forward, but Google has never operated like this. Google will always serve the ad connected to the keyword with the highest ad rank when an exact match keyword is not available.
For similar keywords like +womens +shoes and “womens shoes”, it will likely be driven by the highest bid. For this reason, it is time to say goodbye to BMM or phrase. A simple suggestion would be to keep the match type that has driven the most conversions over the past 30 days, but every account is unique. To ensure queries map to the right keyword over and over, make sure your negative pathing strategy is up to snuff.
While both match types will now operate identically, there is no need to remove BMM terms if there are no phrase match terms in your build. Eventually, you will not be able to create new BMM terms, but as of now, there is no indication that they are in jeopardy of being sunset. Would we bet on that long-term? No, but there is no rush to remove strong keywords and start over.
If you want to clean house, pick your lowest seasonal period to begin, then over the next year, upload phrase match terms into your existing BMM ad groups and campaigns while gradually pausing BMM keywords from your lowest performers on up. Finally, delete these keywords to save space in your account. Your data will persevere in the platform. Again, BMM terms will continue to match as they always have and likely will through at least 2021.
I am not a huge fan of Dynamic Search Ad (DSA) campaigns at scale, but there are valuable use cases for them. The loss of control in phrase can mitigated with more exact match terms in your build. DSA campaigns work by scanning a URL or website and then programmatically matching ads to queries based on content and data Google already has on a site. This can be a great way to mine keywords.
Image: Google
To get started, create a DSA campaign, negate every keyword you have in your main account from that campaign, and let it run for a week or so. Check back and harvest the search queries with the most volume, add them to your account as exact match terms, negate them from the DSA, and repeat. Eventually, negate poorly performing queries from the DSA and continue to monitor. This will help your account to be less reliant on broader match types and be more prescriptive on the performance you expect to see.
Please forgive this former English major for the double negative. There are always ways to take advantage of new changes in digital media. Be a first mover. Even if your account doesn’t have phrase match terms, competitors might and they will be joining incremental and new auctions; places where you have already been. Start new tests, update your assumption sets, and make sure you have contingency plans for the inevitable increase in CPCs that Google is creating by opening up more auctions to more advertisers.
This is also a great time to focus on CRO (conversion rate optimisation) and ad copy testing, because conversion rates might be under pressure with less qualified traffic hitting your site. Do not assume everything will stay the same because your account isn’t changing. This is a dynamic system. Be dynamic.
At the end of the day, Google hasn’t moved the goalposts here. Smart search marketers will take a few weeks to assess the landscape, make some strategic or tactical changes, and continue to drive performance. It will even be a bit nostalgic to have phrase match terms live in accounts again but, as detailed above, there are definitely things that need to be done.
Going forward, as query data becomes harder and harder to parse and automations push results towards the mean and not “up and to the right”, working strategically in that dynamic system will separate the winners from those just taking what Google gives them. Consider this the warmup.