Extracting Valuable Business Insights From P2P Economies
Author: Andrius Palionis, VP Enterprise at Oxylabs
Competition in the ecommerce industry is intense, calling for innovative strategies businesses can use to capture customer interest, maintain loyalty, and generate referrals. Data generated by peer-to-peer (P2P) platforms can provide a competitive edge through unique insights that reveal customer preferences, product quality, and upcoming trends.
P2P platforms exist across nearly every consumer sector, making them a rich data source for mainstream, alternative, and niche categories, as long as no copyrighted or personal data is collected:
- Goods and services marketplaces
- Money-lending platforms
- P2P cryptocurrency exchanges
- Fundraising and crowdfunding platforms
- Online freelancing and talent search platforms
- Ridesharing services
All these platforms generate valuable data for analysis in their specific industry. For this article, however, we’ll be focusing on how professionals in the commercial goods and services sector use web intelligence gathered from P2P marketplaces for their sales and marketing strategies.
Let’s start by reviewing some of the popular platforms and the data they generate:
I imagine most people have heard of Airbnb, a global P2P platform that allows hosts to rent their residences to travelers and long-term guests. The platform is extremely popular with regular travelers, “digital nomads”, and students, with over half a billion guests hosted since its inception. Travel-related businesses typically use Airbnb data to extract pricing trends, occupancy, and geographical insights.
Vinted was one the first P2P platforms that allowed users to exchange or sell clothes, accessories, books, and other personal items, promoting sustainable consumption and circular economy. In 2019, Vinted achieved a ‘unicorn’ status. Currently, it operates in 16 countries. Analyzing data from Vinted, companies can see what brands and products are trending in certain locations and investigate product descriptions to get insights for product improvement and development.
Many products sold online are either branded, mass-produced, generic, or a combination of all these attributes. Etsy provides an alternative shopping environment by connecting consumers with artisans from all over the world that sell handmade, vintage, and specialty items.
Data generated on the search results page includes the product title, short description, star rating, current price, previous price, discount, and time left for the sale (if applicable). Companies selling consumer goods that want to expand their product portfolio can use this data to discover market gaps and predict future trends.
How P2P marketplaces work
P2P platforms are a sum of various technologies that work together to help users conduct buying and selling activities safely and successfully. Different data is generated throughout different stages of the buying cycle; however, some of it must not be scraped due to ethical and legal reasons – for example, data that lies behind logins. Stages of the buying cycle include:
1. User registration
Buyers and sellers typically must create an account by submitting details that include their name, contact information, location, and payment information. This data is highly sensitive and protected by privacy regulations and protocols that defend against data breaches.
2. Listing creation
Once the account is created, users can create listings for their products and services. Some platforms also allow users to create a virtual storefront with seller details, shipping policies, and related information. Public data generated during this stage includes product and service names, descriptions, images, prices, discounts, and additional marketing data.
3. Direct communication
Buyers and sellers interact via direct messaging, chat tools, and other communication interfaces provided on the platform. These channels allow potential buyers to ask questions and negotiate with the seller before making a decision. Communications between buyers and sellers are kept private, and the platform will only get involved to settle disputes between users.
4. Transaction completion and payment
The transaction process is initiated when a buyer decides to proceed with the purchase. Most platforms provide built-in payment systems via services like Paypal or Stripe through methods that include credit and debit cards, bank transfers, digital wallets, cryptocurrency, or cash on delivery.
5. Delivery or service fulfillment
Once the buyer’s payment is finalized, the seller commences delivery of the products or services according to the terms specified in the listing. Some of these details are available for public view on the product or service page.
6. Feedback and reviews
Trust is integral to the success of P2P platforms, which is why users are encouraged to provide user feedback and submit product reviews. Some websites use systems that generate percentage scores based on successful transactions, while others employ a “star” system that assigns value to users, products, and services. Most of this information is publicly available and can be extracted with web scraping tools.
Fictional P2P data scraping use case: Cool Sneaker Co.
Let’s imagine Cool Sneaker Co., a fictional company that wants to disrupt the sneaker market with new designs featuring unique colors and patterns. Before finalizing their product portfolio, they want to be sure that none of their designs has features that are out of style with their target market.
Mass-produced stock that is unsold and out-of-style typically ends up on discount marketplaces. To get an idea of what features to avoid in the sneaker design, the company can use a web scraping API to gather sneaker data containing product titles, descriptions, and images from the largest P2P ecommerce websites.
The first step is to extract the data in compliance with the website’s Terms of Service and existing legislation (for example, images and even titles or descriptions may be subject to copyright while data about the seller may be considered personal information, which should also not be scraped). The images can then be preprocessed and normalized to comparable sizes, resolutions, and formats, and annotated to create a labeled dataset for supervised machine learning.
The second step is to train an ML model on the cleaned and labeled dataset to recognize common colors, patterns, and features. After completing the training and validation process, the model can be applied to the entire dataset to learn the most common colors and patterns for sneakers sold at discount marketplaces.
P2P platforms have completely transformed ecommerce by allowing buyers and sellers to transact without intermediaries. These services also produce data containing critical insights that enable ecommerce companies to optimize marketing strategies, upgrade products and services, and improve the customer service experience.
Web scraping is the leading method for obtaining P2P platform data, but it can be a complex process that requires extensive technical and legal knowledge. However, the benefits of high-quality web intelligence outweigh any temporary challenges because it provides a distinct competitive advantage through unique data-driven insights that enhance decision-making.