AI Negative Decisions Explained
AI Negative Decisions represent the automated actions taken by EVA to exclude underperforming search terms from advertising campaigns.
In advertising systems that rely on keyword or search term targeting, campaigns may appear for a wide variety of search queries. Some of these queries may generate traffic but fail to produce meaningful engagement or conversions. Continuing to serve ads for such queries can result in unnecessary advertising spend.
EVA addresses this by automatically identifying search terms that do not meet defined performance expectations and excluding them through negative keyword decisions.
These actions are recorded in the AI Decisions page so advertisers can review when and why a search term was excluded.
Purpose of negative keyword automation
The goal of AI negative decisions is to improve campaign efficiency by preventing ads from appearing for search terms that are unlikely to generate valuable outcomes.
When a search term consistently underperforms, the system may add it as a negative keyword to stop future impressions or clicks for that term.
By automatically excluding inefficient search terms, EVA helps advertisers:
- reduce wasted advertising spend
- improve targeting precision
- focus advertising budget on higher-performing queries
- maintain campaign efficiency over time
This process allows campaigns to continuously refine their targeting without requiring constant manual keyword management.
When negative decisions occur
Negative keyword decisions are typically triggered when search term performance meets certain conditions defined by optimization rules.
These rules usually follow an IF → THEN logic structure, where specific performance metrics determine whether a search term should be excluded.
For example:
IF a search term generates multiple clicks but produces no sales
THEN add the search term as a negative keyword
Another example may include:
IF click volume exceeds a defined threshold without generating conversions
THEN negate the search term
Once these conditions are satisfied, the system automatically applies the negative keyword and records the action in the AI Decisions table.
Types of negative decisions
Negative decisions generally occur when the system determines that a search term does not provide sufficient performance value.
Examples of situations that may trigger a negative decision include:
- search terms that generate clicks but no sales
- search terms that produce very high advertising cost relative to revenue
- search terms that attract irrelevant traffic
By removing these search terms from campaign targeting, EVA helps ensure that advertising spend remains focused on queries with stronger performance potential.
Where negative decisions appear
When EVA excludes a search term automatically, the action is recorded in the AI Decisions table.
Each entry in the table typically includes information such as:
- the affected campaign or target
- the search term that was negated
- the rule that triggered the decision
- the time the decision was applied
This record allows advertisers to review the negative keyword decisions applied by the automation system.
Monitoring negative keyword decisions
Reviewing AI negative decisions helps advertisers understand how the system is refining campaign targeting.
By analyzing these decisions over time, advertisers can see which search terms are being excluded and evaluate whether the automation logic aligns with their advertising strategy.
If necessary, advertisers can adjust optimization rules or campaign configurations to influence how negative keyword automation behaves.