Bayesian Web Site Testing via Nonuniform Traffic Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current web-site testing methods are costly, time-consuming, and resource-intensive, often disrupting live sites and providing unreliable data due to their invasive nature, which complicates optimization efforts for web-based businesses.
Innovation Solution
A method and system for web-site testing that uses a Bayesian-inference method to nonuniformly distribute web-site accesses among web-page variants, allowing for efficient determination of the most effective web-page variant through minimal HTML-file modifications and third-party testing services, reducing computational and temporal resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If live testing is performed to optimize web site, then web site optimization can be achieved, but testing costs money, time, and resources while potentially creating interruptions and errors for customers
Solution Approach 1:
The patent segments web site traffic into different groups (e.g., via cookies or user identifiers) to assign different web page variants to different segments. This allows parallel testing of multiple variants simultaneously without interfering with each other, reducing the total testing time and resources needed while maintaining optimization reliability.
Solution Approach 2:
The patent applies partial action by testing only the necessary portions of the web site through targeted A/B testing of specific elements rather than complete site redeployments. This minimizes resource consumption and potential disruptions while still achieving meaningful optimization results from the tested elements.
2Ease of manufacture
If uniform distribution of accesses is used among web page variants, then simple implementation is achieved, but determining the most effective variant takes longer and requires more computational resources
Solution Approach 1:
The patent implements dynamic access distribution where the assignment of web page variants to user segments changes over time based on statistical significance thresholds and testing progress. This dynamic approach allows the system to automatically adjust traffic distribution to accelerate identification of effective variants while maintaining implementation simplicity through automated decision-making.
Solution Approach 2:
The patent incorporates feedback mechanisms where performance data from tested variants is continuously analyzed and fed back into the system to determine when statistical significance is achieved. This feedback loop enables automatic termination of testing for inferior variants and reallocation of traffic to promising variants, significantly speeding up the evaluation process without complicating the overall system architecture.
Data Source
AI summary
The current document is directed to methods and systems for testing web sites. In certain implementations of the methods and systems, a testing service collects customer page-access and conversion information on behalf of a web site. The testing service is straightforwardly accessed and configured, through a web-site-based user interface, and is virtually incorporated into the web site by simple HTML-file modifications. A more efficient web-site-testing system nonuniformly distributes web-site accesses among web-page variants in order to more quickly and computationally efficiently determine a most effective web-page variant among a set of tested web-page variants. In certain implementations, nonuniform distribution of web-site accesses among web-page variants is facilitated by a Bayesian-inference method.