Product portraits can be simply understood as labels for massive amounts of data. According to the differences in the attributes of the products, they are divided into different types, and then typical features are extracted from each type, giving them descriptions such as name, price, and category.
Product portraits have important application value in multiple business modules of Amazon’s cross-border e-commerce. For example, when operators want to understand the market profile of a new category, they can use product portraits to analyze the price distribution, review distribution, and sales distribution of the category; when operators want to optimize their own listings, they can also use product portraits to analyze the price strategies and operation strategies used by various competing products in different life cycles of their listings, thereby improving operational efficiency.
In the front-end interface of the Amazon platform, almost all the data built by the product portrait system can be obtained. Taking the exposure results of the search keyword “Sort by Featured” for the category “Clothing, Shoes & Jewelry” on the Amazon platform’s US site as an example, we will explain what product data can be obtained in this search interface.
1. Search exposure page data
The so-called “search exposure page” refers to the front-end page that consumers see after searching for keywords on the Amazon platform.
It should be noted that when obtaining the front-end data of the Amazon platform, select the specific product category after entering the keyword, otherwise the number of products exposed at the front-end may not be complete.
Selecting “All Departments” and “Clothing, Shoes & Jewelry” will result in different numbers of exposed products.
1. Number of exposed products
When the user enters a specific keyword in the Amazon search box and selects a specific category, the front-end will display the exposed products under the keyword, and the number of exposed products and the number of exposed pages will also be displayed.
The data related to the number of exposed products can be used as a reference in both keyword optimization and market analysis. Operators can use this value to determine the volume of product data that can be obtained under a keyword.
Generally speaking, the more “big words” a category has (such as “dress”, “blouse”, “T-shirt” and other words in the clothing industry), the more exposed products there are, and the more accurate the user portrait built through these exposed product data is.
2. Product title data
The product title is the most basic information for each exposed listing.
Getting the product titles of all exposed products under a keyword can help operators determine the changes in high-frequency title vocabulary in different search rankings through word frequency analysis, and thus apply it in new product listings, listing optimization, advertising optimization and other fields.
3. Product image data
Product image The picture data refers to the main picture of each exposed product.
4. Product price data
Product price data refers to the exposure price of each exposed product.
5. Product review data
Product review data consists of two parts, the first part is the review score, and the second part is the number of reviews.
6. Other exposure page information
There is some other exposure information on the search exposure page, such as product promotion information, brand name information, “prime” service mark information, etc. This type of information is not very effective when building a product portrait system. Operators can consider whether to obtain this data based on their own business needs.