In my opinion: dating websites can teach retail about personalised matching, says Neo Technology

Facebooktwittergoogle_plusredditpinterestlinkedinmailFacebooktwittergoogle_plusredditpinterestlinkedinmail

Retail can learn from the highly personalised matching used by dating sites, says Emil Eifrem, CEO of graph leader Neo Technology

Eifrem: graph databases have the power to transform the retail experience

Online dating isn’t just about making relationship connections – it is also very big business and retail can learn much from the way dating websites run.

Online dating has exploded in the last few years. It is estimated to be a £225m business by 2019 in the UK alone, according to Minitel. Globally analysts believe the market is worth around $2bn globally, and growing.

There is nothing revolutionary in the service that match.com, eHarmony, Tinder Zoosk and other giants in this market offer. But they have one thing in common – a significant number of them rely on graph-based technology to make highly sophisticated recommendations to help members find the best matches. Each member believes they are treated as a special individual in this process.

The successful dating sites sell themselves on being a highly personalised service. But underneath, sophisticated technology is at play. Online dating sites are highly skilled at manipulating large sets of connected data. Matches are found by entering users’ profiles, preferences, interactions and connections at the centre of algorithms – quickly identifying the most compatible candidates, so clients don’t have to trawl through thousands of potential profiles to find likeminded souls.

The key to unlocking its success is recommendations. All online dating businesses are underpinned by personalised recommendations, with the most accurate and successful using graph database technology to manage those algorithms. Graph databases differ from traditional business or relational databases in that they specialise in identifying the relationships between very large numbers of data points, and so help users work with data faster and more efficiently.

The online dating site knows exactly what members are looking for and turns up the required results quickly and easily.

Making connections

While the formulae for seeking out compatible partners may vary from site to site, all online dating services are grounded in data and the most accurate matchmaking sites manage and organise this data using graph databases – a technology exploited by digital behemoths such as Amazon and Netflix.

Amazon, for example, is built on its ability to rapidly exploit connections between people and product, and offer “Other people also bought” recommendations.

The power to change retail

Graph databases have the power to transform the retail experience in the same way as online dating sites, allowing companies to match prospective customers with the products or services that are most likely to appeal to them in ever more personalised and immediate ways.

The other differentiating factor is that queries are in real-time or near real-time, since there is no join penalty. Graphs often cross more than three levels deep of relationship while delivering real-time performance. Joining the dots between endless reams of data.

Retailers are starting to understand the competitive edge graph databases can give them. Walmart in the US is already utilising graph database technology to take data harvested from customer purchases at its physical and online stores to the next level.

You may think this is out of the realms of many retail sites. But you would be wrong. Graph databases are no longer the domain of the big players. Database tools and techniques are widely available. While the likes of Google and LinkedIn built their own in-house graph database solutions, off-the-shelf graph databases are now available to any size business.

Effective product recommendation algorithms are fast becoming the de facto standard in online retail — directly affecting revenue streams and enhancing the customer’s shopping experience. In addition, routing recommendations enable companies to cut budgets on routing and delivery, and provide a quicker and more efficient service.

The base line for retailers is the impact graph technology can have on conversion rates, share of the customer’s spending power, repeat business, retailer loyalty and peer-to-peer brand advocacy.

This powerful technology is no longer just the domain of retailers with big IT budgets. The success stories will be retailers, big and small, who seize this opportunity with both hands – taking their customer relationships to another stratosphere.

The author is CEO of Neo Technology, the company behind the world’s leading graph database Neo4j (http://neo4j.com/)

(A Retail Times’ sponsored article)