Common Use Cases for Graph Databases
Microsoft Xbox is one of the most popular gaming platforms in the world, with more than 100 million people using Xbox Live and 18 million subscribed to Game Pass.
All those players generate colossal amounts of data. Microsoft wanted to find patterns in the data to deliver better experiences. Thanks to graph database technology, the company is experiencing benefits as numerous as the alien empires in a “Halo” game.
These include the ability to detect patterns in what users want to keep them playing games longer, intelligently extending special offers and better understanding how groups of users are playing with each other so the overall game experience can be continuously improved.
This is just one example of how graph databases have rapidly become a potent technology spawning new applications all the time, including optimizing business processes, getting a 360-degree view of customers, managing supply chains, improving health-care outcomes, identifying cybersecurity threats and detecting financial fraud.
Gartner predicts that by 2025, graph technologies will be used in a whopping 80% of data and analytics innovations, up from 10% in 2021. Research firm IMACR forecasts that the global graph database market will climb from $1.13 billion in 2021 to $3.78 billion by 2027.
People use graph databases every day and seldom know it. Facebook, Instagram, LinkedIn and Twitter all rely on graph databases and analytics to understand how users relate to each other and to connect users with the right content. Those product recommendations on Amazon — “people who bought this item also bought” — come from a graph analytics query too.
So what makes graph databases interesting?
The story starts with the data explosion all around us. Business leaders constantly ask: What can we do with all this data? And how can we find competitive advantages inside it? They’re under nonstop pressure to get better insights on customer behavior, achieve operational efficiencies and predict changes in the market.
Organizations are gathering data from every possible angle and storing them as independent files in relational databases. These traditional systems are designed to store factors, not analyze data from the point of view of who and where it came from. You may be able to find some isolated facts in the information but miss the unseen relationships within data that enable tackling sophisticated problems with nuance.
While relational database management systems (RDMSs) will remain the operational engine for data transactions and reporting in most enterprises for the foreseeable future, the addition of graph databases opens a whole new window on relationships among data elements.
As Gartner puts it: “Graphs form the foundation of many modern data and analytics capabilities to find relationships between people, places, things, events and locations across diverse data assets.” This allows organizations to “quickly answer complex business questions which require contextual awareness and an understanding of the nature of connections and strengths across multiple entities.”
With graph databases, any type of relationship can be modeled easily, and schemas can change dynamically. However, those fluent in SQL needn’t feel left out, because graph database query languages such as GSQL are simply SQL-adjacent languages augmented with graph capabilities.
And that emphasis on relationships combined with graph technology’s ability to handle enormous quantities of data efficiently have made graph databases an ideal fit for AI and machine learning applications.
Let’s drill down into five use cases where graph databases are proving especially beneficial.
1. A 360-Degree View of Customers
It’s critical for a business to understand how customers are connected to its products and services, as well as to each other. But interactions between companies and their customers or sales prospects tend to be complex, with many touchpoints. Attempts to gain a full view quickly incur many-to-many relationships that, with a relational database, require laborious modeling and loads of cumbersome table joins to produce actionable insights.
Relational databases are good tools for indexing and searching for data, as well as for supporting transactions and performing basic analysis. But graph databases go a step further by connecting across the tables or business entities and identifying hidden relationships and patterns. They are built to understand, explore and analyze the complex relationships in the customer data, allowing data scientists and business users to delve deeper into the data, across all the touchpoints, in real time.
2. Fraud Detection
Fraud is a major problem for financial services firms and many other businesses. Most fraud is perpetrated by rings that aren’t individuals but groups of people collaborating in a complex scheme designed to be hard to track.
The traditional method for flagging fraud is rule-based. For example, if a credit card company sees a really big transaction with a new customer, it denies it. With graph technology coupled with AI/ML, however, you can go as deep as you need to find a sequence or pattern of different interactions, such as a group of otherwise unrelated people sharing the same IP addresses. The deeper you go, the more accurately you can spot real fraud without false positives and detect anomalies that might otherwise be missed.
JPMorgan Chase and Intuit are two examples of prominent financial services companies that have adopted graph databases to enhance their fraud detection efforts.
3. Health Care
UnitedHealth Group (UHG), the world’s largest health-care company by revenue, uses a graph database to help improve quality of care to more than 26 million members, while reducing costs. UHG uses a massive graph database to track more than 120 billion relationships among members, providers, claims, visits, prescriptions, procedures and more.
UHG has developed various applications atop its massive graph database that, among other benefits, provide a consolidated view of member interactions between physicians, pharmacies, clinical labs, health advisers, and UHG itself.
4. Smoother Supply Chains
The name of the game in supply chain management is reducing the risk of operational disruption, increasing site reliability, improving supplier relationship management and managing plant operations cost-effectively.
Many organizations have been able to gather most of the needed data, but their traditional analytic technologies have proven to be too slow, too expensive and generally incapable of analyzing the massive volume of partner, route, transaction and other data stored across various locations, formats and protocols. Most of the traditional supply chain analytics solutions are built on relational databases.
Jaguar Land Rover (JLR) exemplifies how a graph database solution can span the many data silos that needed to be tapped for supply chain analysis and explore the web of relationships among data elements. Some key supply chain-planning queries at the company now take 45 minutes instead of weeks and, more importantly, management can ask questions of data it never could before.
5. Improved Online Retail Operations
Retail e-commerce firms face intense competitive pressure to deliver better customer experiences built on accurate customer details and purchase histories. Early recommendation engines, though breakthroughs at the time, simply looked at a couple of connected data points when making suggestions. They typically used snapshots of days-old data and lacked the real-time modeling and nuanced profiling needed today.
Today’s graph databases offer vast improvement. They can consider an array of possible relationships — such as between customers and payment methods, customers and brands, products and return rates, promotions and sell-through rates, and a whole lot more — in real time to align with the constantly changing profiles of the customer base and produce more engaging and personalized recommendations.
These are just a few of the most common uses for graph databases. The common denominator is this: By putting data in the shape of a graph, organizations can solve complex analytical problems from a relationship perspective, at scale, with speed.
