What Is It?
The Negative Keywords Finder - Shopping helps direct traffic for search queries to the most profitable ad groups in shopping campaigns. It is like an AB test for search queries.
Why Use It?
The tool allows you to:
Eliminate wasted spend from irrelevant or low-performing queries: You can quickly identify search terms that are driving clicks but not delivering results, and block them using exact match negatives to improve efficiency.
Automatically route traffic to the most profitable ad groups: You can ensure that high-intent queries are directed to the ad groups where they perform best, improving overall campaign performance without manual analysis.
Maintain a cleaner and more structured campaign setup: You can avoid internal competition between ad groups by preventing the same query from triggering multiple ad groups inefficiently.
How does it work?
It uses two strategies to recommend negative keywords. Both strategies are completely different and serve different campaign structures.
Duplicate Queries (across ad groups)
This strategy analyzes the performance of the same search queries across different ad groups in a shopping campaign. It finds the ad group in which the query is not performing well and recommends adding it as an exact match negative.
The factors that it uses to determine the losing ad group for the queries are CTR, ROAS, Conversion Rate, and Conversions/impression. This helps direct traffic to the most profitable ad groups.
Low Performing Queries (within ad groups)
This strategy finds queries that are not performing well within an ad group. They may have a lower ROAS than the ad group average or a much higher cost/conversion.
This helps reduce traffic from generic queries. If you’re on a limited budget, these queries can be added as exact match negatives to the ad group.
Filter Criteria
To view suggestions, the search queries need to meet the following filter criteria:
Criteria for duplicate queries (across ad groups)
Impressions should be >= 200
Clicks must be >= 10
Spend for that query must be > 0
The query should either be a winner, or a loser with ROAS < 75% to show up in negative keyword tools' suggestions.
Criteria for low performing queries (within ad groups)
Queries from Only Adgroups with Mean ROAS > 50% are suggested.
Query Impressions should be >= 200
Query Clicks should be >= 10
Query ROAS should be <= 75%
This tool works on the criteria defined above, which means even if you try to change the filter to get a list of search terms that received Impressions < 100, you won't get any results, as there is already a pre-defined filter applied by default.
For your customized requirements, please check out Rule Engine.
Understanding the Results
The table will show a list of search queries that have been analyzed and the ad group in which they are not performing well or losing. These are the ad groups in which they will be added as exact match negatives.
By adding the search terms as exact match negatives in the ad groups they are not performing well in, the tool will help direct traffic to the ad groups they are performing well in or reduce generic traffic.
When you hover on the search term, you'll see an 'i'. Clicking on the 'i' will open a box that will show the ad groups the search term matched to and how it performed. This is to show you more information and give clarity on why the tool is recommending to add the search term as a negative in that particular ad group.
Use Cases
Improving ROAS Across Shopping Campaigns
Problem: Your campaign may have a decent overall ROAS, but certain queries are dragging down performance within specific ad groups. Identifying these manually can be time-consuming and error-prone.
Solution: The tool analyzes performance metrics like ROAS, CTR, and conversion rate to detect underperforming queries. You can act on these insights quickly by adding negatives where needed.
Controlling Traffic from Generic or Irrelevant Queries
Problem: Generic search queries often drive high impressions and clicks but result in poor conversion rates, especially when your budget is limited.
Solution: The tool surfaces generic queries with low performance and recommends excluding them. By adding these as negatives, you limit exposure to low-intent traffic.