Automated Ad Creative Generation and Indexing for E-Commerce
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large e-retailers face challenges in creating and targeting online advertisements effectively due to the time-consuming process of manually generating ad creatives and selecting keywords for their vast product inventories, leading to poorly targeted ads that do not maximize return on investment.
Innovation Solution
An automated system that generates ad creatives and associated information for each product page using a subset of features, creates a page-ad index mapping terms to pages, and selects winning ads through an arbitration process based on search queries, budget constraints, and click-through rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual ad creative authoring and keyword selection is performed for each product, then ad targeting precision and performance are improved, but time consumption and labor costs increase significantly
Solution Approach 1:
The system enables self-service by automatically generating ad creatives and selecting keywords using the product's own web page content. The extraction module retrieves text from product pages, and the generation module creates targeted ads without human intervention, allowing the advertising system to serve itself for large inventories
Solution Approach 2:
The manual mechanical process of ad creation is replaced with an automated computational system. The system uses modules to extract, generate, and evaluate ad creatives algorithmically, substituting human manual work with automated text processing and evaluation mechanisms
2Loss of time
If templated ads are used to reduce manual effort, then time consumption is reduced, but ad quality and relevance to specific products deteriorate
Solution Approach 1:
Instead of using a single generic template for all products, the system generates locally optimized ad creatives tailored to each specific product's web page content. The extraction module captures product-specific text, and the generation module creates unique ad creatives that reflect individual product characteristics, ensuring local quality for each ad
Solution Approach 2:
The system changes the parameters of ad creatives dynamically based on product-specific input. Rather than fixed templates, the generation module varies ad content parameters (text, keywords, descriptions) according to the extracted product information, allowing each ad to have optimized parameters for its specific product
3Device complexity
If general advertisements are created for entire inventory, then device complexity and manual effort are reduced, but ad targeting effectiveness and return on investment worsen
Solution Approach 1:
The system segments the advertising process into distinct automated modules: text extraction from web pages, ad creative generation, keyword selection, and quality evaluation. This segmentation allows each module to handle specific tasks independently, reducing overall system complexity while enabling targeted advertising for each product segment
Solution Approach 2:
The system performs preliminary actions by pre-extracting text from product web pages and pre-generating ad creatives before the advertising campaign launches. The evaluation module also performs preliminary quality assessment, ensuring ads are ready and optimized in advance, which improves ROI without increasing complexity
Data Source
AI summary
Advertiser repository information (e.g., a list of HTML pages) may be used to automatically generate a page-ad and associated ad information for each page using a subset of features of the page. Subsequently, a page-ad index for all pages from different advertisers may be created. After creating a page-ad index, if a client issues a query (e.g., from a search engine), candidate page-ads may be selected using the page-ads index. If necessary (e.g., if there are more candidate page-ads than requested ads), the candidate page-ads may be entered into an arbitration (e.g., auction). Then, the arbitration may determine a set of one or more winning page-ads to be served for rendering.