Online dating is big business. It’s now one of the main ways that singles meet, thanks to its accessibility, scale, reach and published success rates. According to Michael Rosenfeld, a sociologist at Stanford who studies online dating, roughly one of every four straight couples now meets on the Internet and for gay couples, it’s more like two out of every three.
The proposition is wonderfully simple: dating web sites provide access to communities of people searching for the same thing. There’s nothing revolutionary here, since that’s how people have always dated, via their personal or local or professional networks. But what’s distinctive now is the way computers are able to replicate this experience digitally – and at industrial scale – by trawling through vast volumes of data to sift out strong matches and introduce us to compatible individuals.
It’s also something US retailers should be doing. I don’t mean they should be going into the e-dating market, but they need to make all the data they have accumulated from the various multichannel interactions work that much harder for them, in the same way as these online dating firms are doing.
Match products and services just like singles sites do
That’s because all online dating services are built on data, and the most successful matchmaking sites manage and organize this data using graph database technology. Graph databases differ from traditional business (relational) databases in that they specialize in identifying the relationships between very large numbers of data points, at high speed. Online dating back-end technology firm SNAP Interactive, for instance, has nearly 10 billion connections across its database of over a billion unique users, which it has said it can only manage via graph techniques.
But startups like SNAP aren’t the only digital businesses that have to make sense of gigantic amounts of data. Google’s success as a search engine owes much to its ability to rapidly exploit the connections in every Web document, optimizing the relevance and speed of what it serves up to users. Likewise, LinkedIn harnesses real-life relationship networks in such a smooth way that it has dominated the business social network market. Both of these e-pioneers have been using graphs since their first days in business, you’ll be interested to note.
The reason why a graph database is the secret heart of these web giant’s success is that thanks to graphs, connections are identified and followed much more quickly than would be possible using a relational database. A good graph database can query and display numerous connections between people, preferences, personal profile criteria etc., allowing highly targeted content and insights to be served up, at high speed, to interested and soon delighted users.
They are also not shaken when it comes to working at scale and with large datasets; a recommendation engine in a graph database can offer a thousand times performance improvements, despite a thousand times increase in data size, compared to traditional RDBMS.
Super-Personalization
To survive in our increasingly digital world, product and service providers will need not only to understand a person’s past interactions, but be able to instantly combine this knowledge with new interest shown. That’s not just during the customer’s current phone call, text interaction or visit to a bricks and mortar store or online store, but also via their social media activity.
You’re going to want to interrogate this data at lightning speed to serve up compelling recommendations – and offer tailored product offers and promotions. Graph databases are a fantastic facilitator in doing just that, thanks to their ability to effortlessly match historical with live session data.
That’s good news for any retailers trying to make headway in the world of social and across numerous digital channels – as a superior experience leads to happier and more loyal customers, who’ll recommend you to their peers and across social networks. Graph database technology provides an easy way to monitor all of this, meaning US retailers can align themselves with emerging trends – and become ever-more slick and effective at influencing these key influencers.
Perhaps no surprise, then, that retailers are starting to get the graph message. Heavyweights such as Walmart are implementing graph database technology to take information gained from customer purchases at its physical and online stores to the next level. But global sports and athletics giant adidas Group also recently adopted the technology to offer enhanced features such as product recommendations to its audience. In particular, unlike other online retailers that just offer static content on its website, the sports and equipment retailer wanted to personalize content based on user interest, local languages, regional sporting news and market-specific product offerings. Thus whether it’s helping direct a customer to a promotion on their favorite leisurewear, or making connections within a growing digital consumer data set, graphs are already helping adidas deliver super-personalized features to consumers.
The exciting part of all this is that graph database tools and techniques are now available far more widely. While digital pioneers had to build their the first generation of graph data stores, graph platforms are now available to any brand wanting to exploit real-time recommendations to get closer to their customers. It’s thus no longer just the privileged few with vast IT budgets that get to experiment with the next-generation online retail experience, but any size of retailer – even the small independent mid-size retailer.
The challenge is use graph technology to anticipate the needs of your market in ever more creative ways. Even Amazon, which taught us the value in retail of pre-empting what else customers might want to buy, by analysing online sales data, may need to step up here, as newer competitors discover multi-layered, graph-powered data analysis.
Differentiate Yourself in a Crowded Marketspace
It’s no big surprise, then, that graph’s growing in popularity faster than any other type of database – by around 250% last year alone. Gartner thinks 70% of leading companies will be piloting a graph database project of some significant kind by 2018.
The graph database message for retailers out of this is that using graph could mean a huge positive boost for your conversion rates, a chance to up-sell or secure repeat business, customer loyalty and peer-to-peer brand advocacy – while reputation-wise, it’s going to help you be seen as an omni-channel leader, not a follower.
Winning customer hearts is your business, too, don’t you think? So start making graph-powered offers your customers will definitely love.
Emil Eifrem is founder and CEO of Neo Technology.