With the rise of AI, this trend has also reached Amazon operations. A friend of mine said they wanted to find a service provider to adjust their AI advertising, but upon closer inspection, it seems to be more of an automated advertising rather than true AI.
I also tried using my existing tools for testing and settings, and successfully achieved automation for adjusting Amazon advertising. I would like to share the setup method and some parameters with everyone.
Table of Contents#
01 / Implementing Advertising Automation through Amazon Backend
02 / Implementing Advertising Automation through ERP
03 / Rapid Analysis of Advertising Reports using GPTs
04 / Disclaimer
01 / Implementing Advertising Automation through Amazon Backend#
Automatically Download Advertising Report Archives#
Many students who want to review past advertising data find that they can only download data for up to 65 days. However, we can set up automatic downloads of advertising reports in the Amazon Advertising backend. Each automatic ad will have its corresponding record and will be archived in a group. This is very convenient for conducting more comprehensive analysis of advertising data in the future.
In the Amazon Advertising backend, create a report in the Search Advertising Reports section, set the request time to be periodic, and set the end date to the furthest possible time. Once set up, Amazon will automatically download the reports based on our settings within the specified period.
The image below shows the content of our automatic report setup. Whenever we go to this advertising report group, we can download the generated reports at any time.
![[Automatic Download of Advertising Report Archives.png]]
Automatically Adjust Advertising Budget#
In the Amazon backend advertising campaign, under Budget Rules, you can set up automatic rules to adjust the advertising budget based on the conditions we set. There are two types of rule categories: time-based and performance-based. Here, we will focus on the performance-based settings.
- ACOS (Advertising Cost of Sales): Increase the advertising budget by 100% if it is less than or equal to 20%.
- CTR (Click-Through Rate): Increase the advertising budget by 100% if it is greater than or equal to 1%.
- CVR (Conversion Rate): Increase the advertising budget by 100% if it is greater than or equal to 15%.
- ROAS (Return on Advertising Spend): Increase the advertising budget by 100% if it is greater than or equal to 5.
Time-Specific Advertising Placement#
In the Amazon backend advertising campaign, under Advertising Campaign Settings, you can set up automatic CPC adjustments based on the conditions we set to achieve the goal of time-specific advertising placement.
For example, if we set the CPC to $0.5 and set the bid adjustment to 100% from 6-18 o'clock, our bid will be $1 during that time period, allowing us to gain more exposure and clicks during peak hours.
02 / Implementing Advertising Automation through ERP#
All ERPs on the market can perform advertising operations such as keyword bidding, time-specific budgeting, and automatic negative and positive keyword additions. To avoid suspicion of advertising, I won't recommend any specific ones here. I'll just provide some reference values:
- Negative Keywords: Set to not display if there are more than 15 clicks without any conversions, for all ad types except for precise advertising.
- Positive Keywords: Conversion rate greater than 10%.
- Bid Increase: If clicks are less than 10 and impressions are less than 500 within two days.
- Bid Decrease: If conversion rate is less than 10% or ACOS is greater than 50%.
03 / Rapid Analysis of Advertising Reports using GPTs#
Recently, I've been fascinated by AI, so I spent a day training a GPT model for analyzing Amazon advertising based on my understanding of it. Here are the results.
Uploading Search Reports#
![[gpt_ad_analysis_process.jpg]]
Obtaining Optimization Suggestions#
![[gpts_ad_analysis_results.png]]
Currently, it can only suggest positive and negative keywords, CPC adjustments, bid decreases, and advertising placement recommendations.
04 / Disclaimer#
Currently, I have only implemented advertising for beginners and those with a small number of listings. If you are aiming for best-selling and high-quality products, manual intervention is still necessary to achieve better results.
The numerical settings above are specific to my situation and may not apply to all sellers. They are for reference purposes only.