The negative word rule splits your search terms into single words, aggregates their performance, and adds the non-performing words to your ad group as negative phrase keywords.
Navigation: Automation > Negative Word Rule
Use Case
When we analyze the performance of single words, we can often find many irrelevant words causing unnecessary ad spending.
Example: If we sell spatula, our auto campaigns or broad/phrase keywords may trigger the following search terms:
Whenever "toys" or "kids" appears in the search terms, it refers to a different product, is irrelevant to the shopper's buying intent, and has 0 sales.
We found other examples when the SI team analyzed one of our products, Resistance Bands.
1. "bike" with 112 clicks and $47.63 ad spent.
2. "choke" with 129 clicks and $31.05 ad spent.
These 2 words triggered 241 clicks across 194 search terms and cost $78.66 with 0 sales.
If you look at the list of search terms containing these 2 words, you can see that they are unrelated to Resistance Bands.
How do you identify non-performing words?
Step 1: Go to Ads Insights > Main Words.
Step 2: Select the "Unique Word with 0 Sales" filter to find words with at least 25 clicks but 0 orders.
Note: You can customize the filters to change the criteria.
How does the algorithm work?
1. The system takes 60 days of customer search terms, skipping the past 2 days to allow Amazon data to be updated.
Note: You can analyze up to 180 days of historical data and customize the number of days to skip.
Learn how the Negative Word Rule analyzes historical data: How Do Algorithms Analyze Data?
2. The system aggregates every word in the search terms for each product to identify the poor-performing words.
3. The system finds the product's ad groups with the assigned rule.
4. The system adds negative phrase keywords to ad groups with the assigned rule.
Note: If the word appears in your title, the algorithm will not negate the singular/plural form of the word.