The problem with Google Ads broad match no one talks about

In a few days, at the 2024 Google Marketing Live conference, Google will explain how its latest AI developments will unlock new potential for our campaigns.

Contact Us

Google Ads advertisers leverage Google’s technology to reach more users, achieve more conversions, and grow profitably. We have also learned that their features can include major flaws (hello, Performance Max! 👋).

One of Google’s oldest such features is broad match. Google pushes hard for advertisers to use it exclusively: broad match is the default match type, it is featured heavily in their automated recommendations, and Google reps constantly encourage advertisers to ‘upgrade’ keywords to broad match.

In Google’s own words:

(Broad match) helps your ads reach a wider audience without requiring you to build an extensive keyword list. For example, let’s say you have the broad match keyword car window repair. This keyword can reach searches such as automobile glass replacement, even though it doesn't include any of the words as your keyword because they relate to the query.

It seems like a great solution to serve more relevant ad impressions, right? 

At Equeco, we examined whether broad match associations are as relevant as Google’s example of ‘car window repair’ with ‘automobile glass replacement’ suggests.

We analyzed hundreds of search term associations

We manually reviewed hundreds of search term and keyword associations for three accounts and grouped them into three categories:

Ads serving on suboptimal associations have a detrimental effect on ad copy and conversion rate performance. If a user is searching for flights from LAX to NYC and the advertiser has developed keywords, ads, and a landing page for that intent, Google should respect that.

Broad match raises serious quality concerns

An analysis of hundreds of search terms and keyword associations found that only 75% of the search terms had been served on the best possible keyword. 

For the remaining 25% of search terms, there was another keyword better suited to it.

How wrong can Google go? Consider these real-life examples:

One hotel advertiser saw their generic brand term serve thousands of impressions on hotel-specific keywords. This was a major issue because each hotel has its budget constraints.

One yoga retreat advertiser had their ads served for the homonym ‘Bonaparte's retreat’ and the non-commercial term ‘retreat meaning’.

A campaign experiment comparing broad to phrase match campaigns for a real estate advertiser yielded CPAs 78% higher on broad match.

Based on these results, broad match sounds terrible. Should we just avoid it entirely?

Equeco’s methodology

Instead of relying on broad match to target as many queries as possible, Equeco builds and curates extensive lists of highly-targeted phrase or exact match keywords. 

This approach is difficult and has many potential pitfalls, which is why it is a key feature of Equeco technology.

We monitor our search query reports to ensure the search terms are relevant and update negative lists accordingly.

We do occasionally implement successful broad-match experiments for advertisers that meet the following criteria:

  • A large inventory of products that can meet most user intents in the advertiser’s vertical
  • An appetite for growth and a high tolerance for far-flung associations

In conclusion, broad match is a good starting point for advertisers who want to cast a wide net but don’t have the resources to optimize their campaigns. However, keyword research and optimization yield stronger results for those seeking greater campaign precision and profitability.