[Paper Review] Crowd-assisted Search for Price Discrimination in E-Commerce: First results

Jakub Mikians, László Gyarmati+2 more|arXiv (Cornell University)|Jul 17, 2013

Consumer Market Behavior and PricingBusiness, Management and Accounting4 references21 citations

TL;DR

This paper presents a crowd-sourced approach using a browser extension ($heriff) to detect price discrimination in e-commerce, revealing that prices for the same product vary by 10%-30% across different locations and configurations, with extreme cases reaching up to ×2. The study identifies consistent price variations across multiple retailers, including niche and mainstream sites, independent of currency, shipping, or tax differences, suggesting systemic personalized pricing based on user profiles.

ABSTRACT

After years of speculation, price discrimination in e-commerce driven by the personal information that users leave (involuntarily) online, has started attracting the attention of privacy researchers, regulators, and the press. In our previous work we demonstrated instances of products whose prices varied online depending on the location and the characteristics of perspective online buyers. In an effort to scale up our study we have turned to crowd-sourcing. Using a browser extension we have collected the prices obtained by an initial set of 340 test users as they surf the web for products of their interest. This initial dataset has permitted us to identify a set of online stores where price variation is more pronounced. We have focused on this subset, and performed a systematic crawl of their products and logged the prices obtained from different vantage points and browser configurations. By analyzing this dataset we see that there exist several retailers that return prices for the same product that vary by 10%-30% whereas there also exist isolated cases that may vary up to a multiplicative factor, e.g., x2. To the best of our efforts we could not attribute the observed price gaps to currency, shipping, or taxation differences.

Motivation & Objective

• To scale up the detection of price discrimination in e-commerce beyond isolated case studies.
• To investigate whether price variations are persistent, reproducible, and attributable to user-specific factors like location or browsing behavior.
• To determine the magnitude and frequency of price differences across diverse retailers and product categories.
• To rule out alternative explanations such as currency, shipping, taxation, or A/B testing for observed price variations.
• To explore the role of third-party trackers and user personal information in enabling dynamic pricing.

Proposed method

• Deployed a browser extension ($heriff) to collect real-time pricing data from 340 crowd-sourced users across diverse browsing sessions.
• Used the crowd-sourced data to identify a subset of 15 retailers with pronounced price variations for systematic crawling.
• Performed large-scale web crawls from multiple vantage points (different locations and browser configurations) to log prices for the same products.
• Applied template-specific price extraction techniques per retailer to ensure accurate price parsing despite varying web layouts.
• Controlled for confounding factors such as timing differences, A/B testing, and currency/format variations to isolate potential price discrimination.
• Analyzed the impact of user login status and browsing history on pricing, particularly on Amazon for Kindle ebooks.

Experimental results

Research questions

• RQ1 Are price variations in e-commerce persistent, reproducible, and observable across different users and locations?
• RQ2 What is the frequency and magnitude of price variations across different retailers and product categories?
• RQ3 Can observed price variations be attributed to price discrimination, or are they explainable by factors like currency, shipping, or taxation?
• RQ4 Does user-specific information such as location, browsing history, or login status influence the prices displayed?
• RQ5 To what extent do third-party trackers (e.g., Google, Facebook) enable or facilitate personalized pricing?

Key findings

• Price variations of 10%-30% were consistently observed across multiple retailers, with some extreme cases reaching up to a ×2 price difference for the same product.
• The most significant price variations occurred for cheaper products, with some experiencing up to ×3 variation, while more expensive items showed lower variation (up to ×1.5).
• Finland consistently showed higher prices than other locations, with the price in Finland being the highest in 90% of the analyzed retailers, suggesting a regional pricing pattern.
• No consistent price differences were found based on user login status or browsing history for most retailers, though minor variations were observed on Amazon for Kindle ebooks.
• Third-party trackers were highly prevalent: Google Analytics on 95% of retailers, DoubleClick on 65%, and social media widgets (Facebook, Pinterest, Twitter) on 40%-80% of sites, indicating potential data collection pathways for personalized pricing.
• The study ruled out currency, shipping, and tax differences as primary causes of observed price gaps, indicating that the variations are likely due to personalized pricing strategies.