AI Integrates into Cross-border E-commerce: How to Achieve Intelligent Advancement in Amazon Advertising Operations?

[Original Source] Ebrun

AI, without a doubt, is currently one of the hottest topics. Over the past year, we have witnessed the rapid development of OpenAI from ChatGTP to Sora. Many industries and companies have elevated the embrace and effective utilization of AI to a higher strategic priority level.

Although it is unclear what the future will be like, there is no doubt that it is a huge force that can be compared with the inventions of steam engines, electricity, and computers—this is the unanimous conclusion.

In the field of cross-border e-commerce in China, whether it is platforms, service providers, or merchants, how to take advantage of the trend is an important issue. Looking at the AI application achievements that have been verified so far, they mainly revolve around product selection decisions, content generation, advertising placement, and customer service. Especially in the field of advertising placement, with the help of AI-driven data technology, merchants can achieve an upgrade from efficiency improvement to value creation, which can be said to be an important breakthrough for creating additional value.

However, issues such as the underlying logic of AI algorithms, functional boundaries, application methods, data security, and how to integrate with existing processes hinder more merchants from using AI tools to solve operational problems. To address this, SparkXGlobal's smart advertising placement platform —Xmars, organized an in-depth seminar with the theme of "Amazon AI Advertising Operations," gathering many operational experts to exchange Amazon advertising trends, pain points for merchants' operations, the underlying logic of AI advertising operations, and practical experiences in AI advertising operations.

01

AI Seizes the New Heights of Advertising Operations

According to Shen Junhua, Senior Partner Development Manager of Amazon Ads, with the increase in ad inventory and diversified seller demands, and the increasingly rich solutions, Amazon Ads in the global market welcomed strong growth in 2023. In particular, the Chinese market has also grown rapidly, with the number of active Chinese advertisers in 2023 being more than 13 times that in 2017.

At the same time, in 2024, Amazon advertising will embrace five major trends: the application of generative AI, such as the automated creation of ad materials to reduce production costs; the use of streaming TV ads to reach a wider audience and enhance brand awareness; Clean Room data fusion to enhance ad precision and conversion rates; machine learning to improve relevance, accurate targeting, and enhance ad effectiveness; interactive ads to increase user engagement and improve brand experience.

However, in the context of intensified global competition, merchants also face challenges such as complex marketing strategies, the need to comprehensively use multiple advertising products and strategies, and the difficulty in proving the value of marketing. Specifically at the advertising operations level, there are widespread pain points such as a lack of target management capabilities, a lack of effective placement strategies, a lack of systematic data analysis and decision optimization capabilities, and insufficient energy to execute strategies and track ad performance.

AI technology's advances are like a beam of light shining into the marketing field, solving many problems. For example, AI can better judge the budget to use at the next time point based on business or ad placement data, and optimize or place ads based on existing business objectives. In addition, combined with AIGC content, it can achieve more personalized and real-time landing page copy and ad creative; by optimizing marketing data, it can convey more accurate information to consumers, thus forming a complete closed loop from insight to action, making ad placement more efficient.

"Even with SOP, the performance levels of different operations (personnel) vary greatly, and it is particularly important to use AI to raise the average level."

"Amazon advertising operations is a very complex knowledge system; new sellers need to quickly get started using tools, and veteran sellers need tools to do repetitive and low-value work."

"From focusing on Acos value, to focusing on ad structure, and now focusing on data models, the focus on indicator dimensions is increasing. Companies need to iterate on advertising operations ideas through contact with market talent and technological trends."

Several merchants and operational experts at the seminar expressed positive opinions about the value of AI in Amazon advertising operations. They have always believed that, at present, running data through AI to verify and optimize ad operations is a way to greatly improve efficiency.

Taking Xmars, an AI advertising optimization SaaS for Amazon, as an example, in Amazon advertising operations, what AI does is similar to daily operational tasks, or rather, copying and extending human operations. Its scope of action mainly involves bidding optimization, budget optimization, targeted optimization, and ad structure optimization.

02

Understanding AI Advertising Operations through Xmars

Ling Chen, Chief Solution Officer of SparkXGlobal, pointed out in the sharing session that since 2017, AI has entered a period of rapid development, accompanied by the birth of the concepts of "machine learning" and "deep learning." The core of the former is to allow machines to learn automatically and find patterns, different from systems that execute according to artificial rules; the latter requires a large amount of labeled data to train models. The principle of Xmars is to use machine learning to optimize Amazon ad placement and find the best ad placement strategies.

Kun, head of the Xmars data science team, introduced how Xmars optimizes Amazon advertising based on AI. In summary, data and algorithms are the two core elements of Xmars' operation. By collecting and processing a large amount of data, and using steps such as machine learning, predictive modeling, and expert system adjustments, Xmars achieves precise bidding, targeting, budgeting, and ad structure optimization.

Specifically, the underlying operation of Xmars includes three levels:

1. Providing data analysis and decision making for AI optimization with "fast and comprehensive" data. Xmars has comprehensive data sources, including Amazon-related data, Xmars exclusive data (such as product information, keyword information, industry information, etc.), and third-party data (such as customer CRM or other proprietary data), and can access hourly ad data and business result data to make predictions and judgments.

2. Using machine learning models to achieve ad optimization. Xmars can use advanced algorithms (such as 20 prediction algorithms and an "expert system" to supplement adjustment logic for complex scenarios) and fast computing power to predict user ad conversion rates, and automatically provide bidding and budget optimization strategies based on this.

3. Achieving better and more personalized operational effects through the collaboration of humans and AI. In terms of ad creation, merchants set growth-driving, stable order, and event-driven goals, and the AI intelligent recommendation engine realizes intelligent bidding, budgeting, as well as keyword and competitive targeting recommendations; for ad optimization, the AI intelligent optimization engine can automatically optimize bidding, targeting, and budget adjustment based on goals for existing ads and ad campaigns.

In addition, explainable AI and personalized AI are major features of Xmars. Explainable AI helps merchants intuitively understand the reasons behind AI adjustments by providing decision-supporting data analysis; personalized AI, through merchant's independent parameter adjustments (such as adjusting AI execution frequency, adjustment amplitude, data review window, etc., to set the degree of AI "aggressive" and "conservative"), gradually trains and interacts, making AI more and more in line with the merchant's personalized needs.

03

How to Understand and Apply AI Correctly?

Summarizing and accumulating the actions and processes that merchants carry out in Amazon advertising operations in their daily work, to achieve batch operation and rule-based automation—this method does not involve AI. It is simply a process of machine automation, that is, using simple mechanical execution to improve efficiency. However, AI makes further judgments based on the merchant's business objectives and marketing objectives, and actively creates, optimizes, and improves ads in a full closed loop.

In other words, AI can be seen as a super partner and collaborator for merchants in operations. The work done by AI is very similar to that done by people, but AI, with a large accumulation and complete algorithm logic behind it, ensures that each action is more accurate, timely, and in line with future business objectives than a person.

However, when actively engaging with cutting-edge technology, strengthening AI knowledge learning, and selecting suitable AI tools to assist in ad operations based on their own business needs, merchants also need to understand the applications of AI correctly: firstly, a large amount of data is essential for the effectiveness of AI tools, so sufficient data is the foundation of everything. Secondly, one should not expect quick results, as AI takes time to learn and optimize; it is a gradual adjustment process.

As Daniel Dong, founder of the Yifeng E-commerce Academy, shared at the seminar: Using AI well requires a process, and it is most important to manage expectations, spend time understanding it, and patiently wait for its effects.

Daniel Dong shared from his own experience the three stages he went through after introducing Xmars: at first, expectations were high, and the data tended to be positive, but the results did not meet expectations; later, they began to adjust expectations, lower them, and reduce the frequency of manual intervention, and after a period of data accumulation, the effects gradually approached the target; now, they have set reasonable goals, gradually reduced ACoS in an organized manner, and after about 20 days, ACoS successfully decreased from 30%+ to 20%+, ultimately achieving an ideal synergistic state with Xmars.

According to his introduction, the period for achieving results using Xmars is 14-25 days, with a success rate of approximately 80%. In its view, AI empowerment is a powerful aid for Amazon's daily operations: "Currently, we use XMars to accomplish many tasks that cannot be done manually, such as time-based pricing adjustments, time-based budgeting, and ranking locking. This allows our operations to break free from trivial tasks and engage in more strategic work."