A phenomenon of the Internet age, online dating has taken off in a way that previous match-making services – such as personal ads and lonely hearts columns – never managed. In fact, online dating is now one of the most popular ways to meet a new partner, with over 1,400 sites in the UK alone, catering for people from all walks of life and interests.

Essentially, cyber-dating is all about access to a group of matching singles – a community. And that’s where services such as Match.com, eHarmony and Zoosk excel, giving singletons access to a much wider set of prospects than was possible before. The same is true of niche sites such as Uniform Dating, Dance Passions or Farmers Only, which, because so much more targeted, gather together very specific sub-groups of people seeking companionship.

There is nothing revolutionary here. Before online dating or personal ads, the traditional way of meeting potential partners has always been through extending your current relationships – usually through friends, family, colleagues and neighbours. Really, that’s all online dating is doing – putting people’s interactions and connections at the centre of its algorithms to help you search for interests and attitudes and life experiences that you have in common with others.

What Retail Can Learn

While the formulae for love may vary based on the dating site, they are all underpinned by data. This is where graph databases make a difference, with digital leaders like Google and LinkedIn owing their breakthrough to having adopted this revolutionary approach to data very early on.

Graph databases differ from relational databases – as the vast majority of everyday business databases are called – as they specialise in identifying the relationships between multiple data points. Thus Google was able to exploit the connections in every Web document and build a connected dataset to rank search result, getting back substantially better search results: a fundamental factor in its meteoric rise.

LinkedIn, meanwhile, digitises real-life relationship networks, common business contacts and friends-of-friends in a slick way that has given it total domination of the business social market space. Those relationships can be spotted and followed by the code much more quickly than in a relational database because those ‘relationships’ (in the graph database world, ‘joins’) don’t need to be created in order to run a query. This means much improved query time, supporting speedier transactions. The graph database can also query and display a huge volume of connections between people, preferences, personal profile criteria, and so on.

And although not dealing with two human connections, managers in a very different industry to online dating – retail – should be able to see the potential in matching up prospective customers with their ‘ideal’ products or services.

Using those connections is something retailers can hugely benefit from as well, it turns out.

What’s helping is a huge ‘democratising’ of these technologies. While social web pioneers and market giants like Google and LinkedIn had to build their own in-house data stores from scratch, off-the-shelf graph databases are now available to any business wanting to make the most of real-time recommendations.

No surprise then that graph databases are growing in popularity faster than any other type of database – by around 250% last year alone. Forrester Research estimates that one in four enterprises will be using such technology by 2017.

The real point is that while at one time it was hard to see how ordinary firms would ever be able to emulate the resource-rich web giants when it comes to data architectures, that is exactly what both dating sites and retailers are starting to do.

The Community Approach

Online dating also owes much of its success to the way graphs can analyse even the most complex relationships, looking not only at location and personal details but also passions, hobbies and attitudes to identify potential matches. It turns out that understanding that richer side of a person’s life is a far better indicator of a potential love-match than a purely statistical analysis of more abstract tastes and interests.

Retailers are already starting to get that message, too. Organisations such as Walmart – the largest company in the planet – are implementing graph database technology to analyse the wealth of peer-to-peer information it’s getting from analysing customer purchases from both their physical and online stores. Amazon has taught us that to be a successful online business, recommendations are key to success; it’s a fantastically useful tool to increase the bottom line. You can look at online transections as tables or isolated business tables, but if you can look at them as a graph then you can do easy collaborative filtering-type recommendations very quickly, something trivially easy to do in graph databases – but next to impossible in relational databases.

Moreover, looking at online shoppers’ behaviour and the relationship between customers and the products they purchase has allowed smart retailers to improve real-time product recommendation service, giving customers a level of personalised service that self-service stores have rarely been able to offer. Expect finely-tuned recommendations becoming more and more of the norm and the Amazon one-size-fits-all suggestion getting left behind as a result.

To do that, real-time recommendations will need you to understand the customer’s past purchases, quickly query this data, and match the customer to the bigger pattern that is their closest match in your database, both in their social network and in buying patterns. Making on-the-fly real-time recommendations also requires the ability to instantly capture any new interests shown in the customer’s current visit.

The good news is that, again, matching historical and session data is trivial for graph databases. It’s got to be a win-win. Better service leads to happier and more loyal customers who will recommend your organisation within their community.

That’s why graph is becoming more and more of a buzzword for the entire e-commerce universe. According to Gartner, for data-driven operations and decisions Graph Databases are “possibly the single most effective competitive differentiator”. That’s great for retailers adopting the approach, who can also benefit from adding in a social element to all this; by looking at customers’ connections and social media activity retailers can quickly identify those likely to speak about a brand and recommend services to others.

Imagine being able to easily ask things like, “Which customers of mine are mentioning my brand on social media and via which channels?”, “Which have most social influence amongst their networks and like to write reviews?” and then being able to correlate all that great insight with data like, “Now tell me which of these digitally active customers are buying most of our products?” Patently, being able to do all this will give your business immensely valuable insight into potential trends. The answers will also tell you who are the most appropriate individuals that you can engage with and nurture into brand advocates, or track as trendsetters.

So let’s put all this together. What have we learned? That graph databases offer a new way of helping customers, connecting with them on a rich, personalised level. Ultimately, they can also improve the conversion rate of purchasing and enhance customer loyalty, adding money to the bottom line and providing real competitive advantage. And that’s got to be a match made in Heaven, surely?

Emil Eifrem

Emil Eifrem


Emil Eifrem, CEO, Neo4j