A digital marketing manager has upgraded their keywords to broad match. Which behavior should they build into their regular cadence of optimization to guide machine learning?

When a digital marketing manager upgrades to broad match, the primary goal is to leverage machine learning to discover new, relevant search queries that might have been missed by more restrictive match types. To effectively guide this machine learning and optimize performance, the system needs flexibility and a clear signal of what is relevant.

Analysis of Correct Answer(s)

• Remove negative keywords that may block relevant traffic.
  • Explanation: When using broad match, the machine learning algorithm explores a wider range of queries. If existing negative keywords are too restrictive or inadvertently blocking queries that are actually relevant variations or synonyms, they can hinder the learning process. By carefully reviewing and removing negative keywords that might be too broad or are blocking valuable, relevant traffic, the manager allows the machine learning algorithm to explore these new paths, learn from their performance, and expand reach effectively. This provides the necessary data for the algorithm to optimize, identifying what works and what doesn't within the expanded scope of broad match.

Analysis of Incorrect Options

• Create additional phrase and exact match keywords to improve reach.

  • Explanation: While phrase and exact match keywords are valuable for control and specific targeting, the manager just transitioned to broad match, indicating a desire to use machine learning for discovery. Adding more specific keywords immediately after going broad match can counteract the goal of broad match, which is to allow the system to explore and find new, relevant queries through its own learning. It also creates more internal competition rather than guiding the broad match's exploration. The focus should be on guiding broad match, not immediately overriding its function with manual, specific targeting.

• Add Customer Match lists with data older than 90 days.

  • Explanation: Customer Match lists are excellent for targeting specific known audiences. However, data older than 90 days may be stale and less effective for current campaign performance. More importantly, Customer Match is an audience targeting strategy; it does not directly guide the machine learning algorithm for keyword optimization or broad match discovery. While valuable for overall campaign performance, it's not the primary lever for guiding broad match keyword exploration.

• Dismiss recommendations to add relevant new keywords.

  • Explanation: Dismissing recommendations to add relevant new keywords directly contradicts the objective of using broad match to leverage machine learning for discovery. These recommendations are often generated by the platform's machine learning based on observed user behavior and potential opportunities. Dismissing them would prevent the account from exploring potentially valuable search queries, thereby hindering the machine learning process and limiting the potential for broad match to find new, effective traffic. The goal is to provide more data and signals, not to restrict them.