We have entered an era where e-Commerce rules retail. Consider how reports project online sales to hit more than $4 trillion by 2020, representing 14.6% of total retail spending worldwide (source: eMarketer). Over the past few years, online shopping has transformed how we buy, bringing in a new “Age of the Consumer.” Today, shoppers have access to more information than ever around products and brands, all informing their purchasing decisions. They also are taking the lead in their own online shopping experiences, with the global marketplace available at their fingertips.
In turn, e-Commerce has embraced AI-powered assistants, recommendation engines and even automated platforms to help consumers consider what to evaluate and buy. Insights generated from AI platforms provide tremendous value, and can potentially drive further revenues for companies, as well as provide shoppers with a better, more customized customer experience.
Traditional brick-and-mortar sales relied, and continue to rely, on establishing a connection with the customer in the form of discerning and evaluating their needs, making recommendations and walking them through the sales process. Connections matter even more so when it comes to online sales, as the most successful retailers today offer a customer-centric e-Commerce platform. This involves a consistent multipath purchase experience, achieved through a 360-degree view of each and every customer. Being able to translate customer data and activity into actionable intelligence to support the customer during their shopping experience with product recommendations and more can lead to increased conversions and higher units per transaction.
The key to successfully leveraging this data is having it and being able to analyze it effectively in the first place, not to mention being able to do so in real time, during a customer’s live shopping session. Most e-Commerce retailers are able to gather the data, but are finding that traditional analytic solutions are missing the mark to achieve this. Historically, solutions have been too slow or expensive, and often are incapable of deriving sophisticated insight from massive amounts of customer, transaction and external data.
Real-time offers — be it products on hand, conditional promotions or free shipping to save a sale — require knowing your customers, and knowing them well. This means e-Commerce sites need to collect and leverage consumer segmentation data as quickly as possible to provide targeted recommendations and customer service. The result is being able to offer a more personalized shopping experience that customers enjoy, engage with and consistently come back to for their needs.
Delivering A Better Online Shopping Experience…And Driving Revenue
Achieving this is no easy feat, especially using traditional databases that tend to store data in tables. That’s why more and more retailers are turning to the power of graph database technology to support real-time analytics and more. For retail, graph databases are the ideal choice for an industry where connections are of the utmost essence. This is because graph databases are designed from the ground up to treat relationships as first-class citizens, providing a structure natively embracing the relationships between data. This ensures better storing, processing and querying of connections than ever before.
For example, consider a simple personalized recommendation such as “customers who liked what you liked also bought these items.” Starting from a person, a query made in a graph database first identifies items viewed/liked/bought. Second, it finds other people who have viewed/liked/bought those items. Third, it identifies additional items bought by those people.
Person → Product → (other)Persons → (other) Products
To power such queries, over the past few years graph databases have seen a major uptick in enterprise adoption, particularly by retailers powering e-Commerce sites. While the first generation of graph database solutions have been helpful in addressing the e-Commerce data problem, they are not without their limitations. Query speeds and analysis of data are relatively slow, with graph engines only able to support traversals (the process of visiting, or checking and/or updating, points of data) to a certain extent. Deeper traversals allow you to glean better insights from data.
Empowering E-Commerce With Real-Time Deep Link Analytics
Recently, we have seen the next-phase in the graph database evolution, with technology fulfilling the needs of e-Commerce by providing Deep Link Analytics. This enables customer intelligence in real time, along with powerful relationship analysis. With these real-time capabilities, e-Commerce sites can quickly synthesize and make sense of customer behavior. The result is the capture of key Business Moments, transient opportunities where people, businesses, data and “things” work together dynamically to create value used to personalize the customer experience, which leads to more transactions.
This is accomplished as new graph database technology supports 10+ hops (or traversals of data points as demonstrated above), unlike being limited to just two as first-generation solutions are. Each hop adds more intelligence about an individual user’s shopping history — whether it is their first visit or they are a repeat customer. Analyzing the data/product/customer location/weather based on location/recommended products requires the capability to traverse the data to present a recommendation in real time.
The query above requires three hops in real time, so it is beyond the two-hop limitation of current-generation graph technology on larger data sets. Adding in another relationship easily extends the query to four or more hops. This data continuously feeds the AI and machine learning algorithms designed to bring a better online shopping experience.
Deep Link Analytics allow e-Commerce companies to be able traverse these data, creating smarter AI and machine learning to deliver competitive advantages, close more deals and build customer loyalty. It all comes down to being able to derive better insight by considering more connections among your data to adequately address a shopper’s online behavior and preferences.
Deep Link Analytics For Product Recommendations
Let’s consider an example of Deep Link Analytics being used by a retailer today. A major global e-Commerce platform uses Deep Link Analytics to model its vast product catalog, map its consumer entity data and to perform real-time analytics over its complex and colossal amounts of data. This leads to real-time recommendations that are personalized for each shopper, driving sales and revenue.
At a technical level, the solution has increased query speeds by 100X compared to the company’s previous in-house solution, with times under 100 milliseconds. It also has saved memory usage by 10X, achieved through efficient graph based encoding and compression. As fewer machines are needed, this also has helped cut hardware management and maintenance costs.
Data scientists and machine learning experts are able to do more, faster, as they are able to use a single massive graph integrating all the company’s connected data. The business derives insight from its massive amounts of data to support better decision making and improving time to market. The graph database platform also provides the speed and scalability needed to make the most data relationships for competitive advantage.
The Customer Connection
Retailers are finding that connecting with consumers online is just as important as connecting with them in person when they shop in a brick-and-mortar location. To achieve this, it has become more and more relevant for e-Commerce to find ways to scan through huge amounts of data and to make sense of it.
Today, it’s more possible to do this than ever before, especially using a graph database where connections among data are at the forefront. As graph databases mature, they are powering deeper insight by supporting queries over more connected data. Customers enjoy a better shopping experience, while retailers benefit from happier and more engaged shoppers along with additional sales and revenues.
Dr. Yu Xu is the founder and CEO of TigerGraph, the world’s first native parallel graph database. Dr. Xu received his Ph.D in Computer Science and Engineering from the University of California San Diego. He is an expert in Big Data and parallel database systems and has 26 patents in parallel data management and optimization. Prior to founding TigerGraph, Dr. Xu worked on Twitter’s data infrastructure for massive data analytics. Before that, he worked as Teradata’s Hadoop architect where he led the company’s Big Data initiatives.