Overview
The new Keyword and Targeting Actions Page (Beta) from Teikametrics is designed to streamline how you manage keyword and target additions and negations. This update introduces powerful AI Smart Filter recommendations, customizable filters, and insightful performance tracking—all in one place. The enhancements available in the beta will significantly improve the efficiency and precision of your advertising optimization workflow.
AI Smart Filters
One of the most impactful additions to this release is the introduction of AI Smart Filters. These filters are automatically generated by Teikametrics’ AI models and are tailored to each merchant and action type.
These filters are designed to identify the most promising opportunities based on your historical performance and campaign behavior. You can trust that the recommendations surfaced through these Smart Filters are optimized to drive results.
Each AI Smart Filter varies depending on the ad type (e.g., Sponsored Brands vs. Sponsored Products) and will automatically display any recommendations matching the criteria.
Custom Filters
To further empower advertisers, Teikametrics now supports Custom Filters. These allow you to define your own criteria for keyword and targeting actions based on your business needs.
How Custom Filters Work:
Name Filter: Name your filter and choose whether it should be set to "Review" (available now) or "Automatic" (coming mid-June).
Accepting: Choose to review and accept your recommendations or have them added automatically (automatic coming soon)
Select Ad Type: Choose between Sponsored Products or Sponsored Brands
Targeting Type: This will be dynamic based on the ad type you select but choose between keywords to add, PAT targets to add, or keyword to negate.
Target Intent Type: Choose the target segment you would like to see in your filter - brand terms, competitor terms, generic terms, or a combination of the three.
Note: To use this feature effectively, be sure to input your brand and competitor keywords so the system can classify targets accurately.
Define Criteria: Use performance metrics such as ACOS, clicks, conversions, CTR, and CPC to create specific rules (e.g., ACOS ≤ 50%, Clicks ≥ 5, Conversions ≥ 2).
Apply to Ad Groups: Choose whether to apply your filter to all ad groups or a selected subset.
Once created, your custom filters will automatically start to harvest recommendations based on your criteria. This can take up to one minute. Your new filter will persist until deleted, always searching for the optimal targets to recommend.
AI Value
The AI Value section offers visibility into the impact of AI-driven decisions by showing:
Total attributed sales from keywords you've accepted based on AI recommendations.
Time saved by using AI for keyword addition and negation.
This data helps quantify the performance improvements and operational efficiency gained by adopting AI-powered optimization.
Note: If you're only selling on Walmart, you won’t see keyword negation recommendations. Negation actions are currently only supported for Amazon-connected merchants.
Frequency Recommendations
Teikametrics AI system will evaluate all sources for new recommendations once a week. This allows users time to take action and new recommendations to accumulate data.
If a user updates brand and competitor terms for a merchant this will trigger a refresh of the recommendations which will take up to 24 hours. This refresh will use the new brand and competitor terms to re-tag all targets with the correct target intent.
A 24 hour refresh will also happen if the user edits keyword routing tags for their ad groups. Similarly, the AI system will re-evaluate the destination ad groups for each recommendation based on the new inputs.
Brand and Competitor Terms
Entering your brand and competitor terms in the Recommendations Hub is critical for accurate targeting. These inputs help the AI system understand which keywords are associated with your brand and who your competitors are. With this context, the AI can correctly tag each keyword and route it to the appropriate ad group.
Without this information, all keywords will default to generic, leading to inaccurate placements in your generic ad groups—ultimately resulting in less effective ad performance and suboptimal campaign management.
What’s Not Included in the Beta
While the beta offers powerful capabilities, a few features are not yet available and are scheduled for release in mid-June:
Automation: The ability to automatically apply recommendations that meet your filter criteria without manual review.
Brand and Competitor Classification Callouts: Clear labeling of keywords as Brand, Competitor, or Generic within the review experience.
History Page: A new page providing historical visibility into past keyword recommendations and actions taken.
These features are in development and will further enhance the full release of this functionality.
This beta is currently available to select accounts managed by our Customer Success team. We're excited to get your feedback and continue refining the experience as we approach the broader launch.
If you have questions or need help getting started, please reach out to your Teikametrics representative.