AI Explanation of Keywords Negation: Understand Why AI Negates

1.  Background

In the daily use of the AI negative word function, users are often unable to clearly know the specific basis for the AI to perform the negative word operation. This causes users to have certain confusion about the AI's negative word actions, reducing their trust in the function and their usage experience. To solve this problem, we have launched the Negative Word LLM function, which aims to explain the decision-making basis of the AI's negative word actions, make the AI's negative word actions more transparent, and improve users' usage experience and trust in the function.

2.  Function Introduction

The core function of this iteration is the Negative Word LLM function, which can clearly present the decision-making basis for the AI to perform negative word operations, specifically as follows:

The decision-making basis when the AI performs precise negation and phrase negation.

It supports users to express their agreement or disagreement with the AI's negative word actions. If the user chooses to disagree, they can further select the reason for disagreement (do not recognize the AI's negative word rules, unclear explanation, other), and can input suggestions or ideas about the AI's negative words (optional).

3.  How to use

1. First, enter the AI data overview interface, find the keyword negation topic, and click the [Details] button.

2. Hover the mouse over the row of the negative keyword, and the Negative Word LLM icon will appear.

3. Hover the mouse over this icon and click [Click to view AI negative word explanation] to call up the AI explanation window.

In the AI explanation tab, you can see:

When the AI performs precise negation, it will comprehensively consider the advertising campaign budget based on the performance of the advertising campaign within the lookback window, the set AI goals and personality, and the specific performance of the search term, and finally perform precise negation.

At the same time, it will also consider the operation of other action spaces to cooperate with the negation. For example, for high-potential words in automatic advertisements, if they have been harvested into manual advertisements of the same managed group for more refined bid adjustments, precise negation will be performed in the original automatic advertisements to ensure that the word is only bid and optimized in the manual advertising group.

In addition, if the word has been manually and precisely negated by other ad groups under the same ASIN in the past, this negation operation will be inherited to avoid the waste of costs caused by confirming invalid search terms. 

When the AI performs phrase negation, it will negate the root word based on the conversion and cost of the root word at the parent ASIN level within the lookback window.

Similarly, if phrase negation has been manually performed in other ad groups under the same ASIN in the past, the AI will also learn this operation and perform phrase negation.

4. It supports expressing agreement or disagreement with the AI's negative word actions according to one's own situation:

If you agree, no additional operation is required.

If you disagree, select the corresponding reason for disagreement, and you can input relevant suggestions or ideas as needed.