In my opinion: database techniques can help retailers match the marketing of big brands, says Neo Technology


Eifrem: even small brands can compete

Amazon you know about – but adidas Group and Walmart are also signature brands using graph database techniques to mine data relationships and offer customers a superior experience, says database expert Emil Eifrem

We’re all very familiar with Amazon’s, ‘Customers who bought this like you, also bought this’ recommendations. It’s a brilliant tool; Amazon has shown us the value of winning customer business by providing the most personalised product and service recommendations possible.

Even the smallest retailer knows that to survive in our increasingly digital world, it needs to offer highly personalised product and service recommendations to delight and engage our increasingly demanding global customer. We also all know that firms that don’t meet these demands and which fail to offer this super-personalised experience will pay a penalty on Twitter – and may even go under.

However, the good news is that you, too, can start doing Amazon-style super-personalised customer interactions. The reason: techniques pioneered by Amazon and the other first wave of e-tailers and social web giants like LinkedIn were powered by a technology developed in-house called the graph database.

How do graph databases make a difference? To do personalisation well, you need to look up the customer’s past purchases, query that history then match the customer to the product or promotion that’s the closest match to their interests, ideally based on their social network footprint as much as previous buying patterns. To make the recommendations as sharp and relevant as possible also requires the ability to instantly capture any new interests shown in the current visit. And all this needs to be processed in milliseconds – as the next online temptation is only ever one click away.

In addition, graph databases are able to do a number of things better than the familiar relational database management systems that most organisations use for database tasks. For example, they are able to rapidly match historical data with live session data and secure a millisecond real-time response, as well as identify relationships between very large numbers of data points, so help you work with all manner of product and customer and social data.

They are also very powerful when it comes to working at scale and with large datasets. A recommendation engine in a graph database can offer a staggering thousand times performance improvement, despite a thousand times increase in data size – invaluable when working with the multi-channel sources of data that are exponentially large.

No longer just for the big players

Retailers are implementing graph database technology to make real-time personalised recommendations

Retailers are implementing graph database technology to make real-time personalised recommendations

Large global retailers are starting to use graph databases to compete with the Amazons. The largest of them all, Walmart, is deploying graph database technology to combine information from customer purchases at its physical and online stores in order to make real-time personalised recommendations.

Meanwhile, global sports and athletics giant adidas Group has also adopted the technology to offer enhanced features such as product recommendations to its audience. Unlike other online retailers that offer static content on their website, the retailer wanted to personalise content based on user interests, local languages, regional sporting news and market-specific product offerings. As a result, its internal business users can categorise and search for user trend content across every platform and division of the enterprise – from sources ranging from marketing campaigns, product specifications, contracted athletes and associated teams to sports categories, gender information and more. Whether it’s helping direct a fan to a great piece of their favourite team’s football kit, or making connections within a growing digital consumer data set, graphs are enabling adidas Group to deliver the super-personalised features to consumers we’ve been talking about.

Competitive differentiator

Maybe no surprise, given how big these firms are. But the good news is you needn’t be a big global brand to introduce graph technology, which means it’s giving all sorts of brands – even the smallest operator – an invaluable leg-up in terms of making sense of all this customer data. For example, Gartner predicts that for data-driven operations and decisions, graph databases are now “possibly the single most effective competitive differentiator”.

The attractions of a technology that can provide a 360-degree view of a customer in real time are clear. Don’t miss out on the promise of graph. After all, given its move into retail, you need to be competing with, and indeed outcompeting, Amazon these days – and offering highly sophisticated and personalised recommendations that will keep your customers coming back for more.

The author is co-founder and CEO of Neo Technology, the company behind the world’s leading graph database, Neo4j (

(A Retail Times promotion)