In the last 15 years, eBay grew from a simple website for online auctions to a full-scale e-commerce enterprise that processes petabytes of data to create a better shopping experience. For e-Commerce businesses of this era data mining and machine learning algorithm plays an important role in the following areas –
1. Product Search
When the user searches for a product, how do we find the best results for the user? One factor used in product ranking is user click-through rates or Product sell-through rate. In addition, user behavioural data gives the link from a query, to a product page view, and all the way to the purchase event. Through large-scale data analysis of query logs, we can create graphs between queries and products, and between different products.
We can also mine data to understand user query intent. When a user searches for “Honda Civic”, are they searching for a new car, or just repair parts of the car? Query intent detection comes from understanding the user, other users’ searches, and the semantics of query terms.
2. Product Recommendation and Promotions
Typical recommendation systems are built upon the principle of “collaborative filtering”, where the aggregated choices of similar, past users can be used to provide insights for the current user. Predictive analytics makes this challenge easier by using machine learning to understand a consumer’s behaviour, including the purchase history of that consumer and the performance of different products on the site, to determine relevant recommendations that have a higher probability of generating a sale.
It does the same thing with promotions to identify those that have worked in the past, and then offer the best promotions in real-time based on the consumer’s browsing pattern.
3. Fraud Detection
This is a problem faced by all e-commerce companies. For example, sellers may deliberately list a product in the wrong category to attract user attention, or the item sold is not as the seller described it. On the buy side, all retailers face problems with users using stolen credit cards to make purchases.
Fraud detection involves constant monitoring of online activities, and automatic triggering of internal alarms. Data mining uses statistical analysis and machine learning for the technique of “anomaly detection”.
Detecting seller fraud requires mining data on seller profile, item category, listing price and auction activities. By combining all of this data, we can have a complete picture and fast detection in real time.
4. Business Intelligence
The unique problem facing many e-commerce players is their large and diverse inventory with items ranging from collectible coins to new cars. There is no complete product catalogue that can cover all items sold on their website. How do we know the exact number of “sunglasses” sold on eBay? They can be listed under different categories, with different titles and descriptions, or even offered as part of a bundle with other items.
Inventory intelligence requires the use of data mining to process items and map them to the correct product category. This involves text mining, natural language understanding, and machine learning techniques. Successful inventory classification also helps in providing a better search experience and gives a user the most relevant product.
5. Anticipatory Purchases
The popular audio-recognition app Shazam now has an always-on feature that listens for audio all day long, tags songs that don’t exist in your library, and makes them available for your perusal, saving you the hassle of unlocking your phone, loading the app and nearly causing an accident in your car. With machine learning, Shazam and products like it could intelligently filter your auto-tags to what you’d most likely be interested in based on your existing library. And further than that, with deep neural networks, be able to decide which to go ahead and add to your library, even making a purchase on your behalf.
Amazon recently filed a patent for “anticipatory shipping,” which ships and order before it’s even placed. Amazon’s wealth of order history, search, wish list and click stream data may one day be leveraged this way.
6. Pricing Management
Analysing pricing trends in correlation with sales information to determine the right prices at the right time to maximize revenue and profit. Pricing is managed using a predictive model that looks at historical data for products, sales, customers, and more. Based on this model, the price for a given product and customer can be predicted at any given time. Amazon is a heavy user of predictive pricing, pricing products in real time based on a pricing algorithm that looks at several inputs, like competitor pricing and product pricing trends.
7. Supply Chain Management
Predictive analytics helps understand consumer demand, to effectively manage the overall supply chain process. This includes planning and forecasting, sourcing, fulfilment, delivery, and returns. If a retailer can predict the revenue from a specific product — say in the next month — it results in better inventory management, optimized use of the available warehouse space, better use of cash flow, and avoiding out-of-stock items. Walmart recently acquired Inkiru, a predictive analytics start up with models for supply chain optimization.
This is not an exhaustive list, if you have some more use cases, please share below.
Rohit Yadav is a customer experience evangelist helping companies identify and make the best use of their key performance indicators and generate insights to improve their customer experience. Rohit is a regular writer on technology, analytics and customer centricity for various leading forums like Analytics India Magazine, KDnuggets, Data Science Central, CX Journey, MyCustomer.com & CustomerThink.com. Follow Rohit on Twitter @roityadav