Macquarie banks on customer appetite for a ‘Spotify-like’ experience

Jessica Twentyman Profile picture for user jtwentyman September 29, 2016
Australian financial services provider launches new platform for consumer banking that relies on technology from Datastax to provide natural language search and machine learning capabilities.

Ben Perham

Financial services company Macquarie Group is better known as an investment bank than as a provider of retail banking services, a market in which it lags behind four major competitors on its home turf of Australia: ANZ, Westpac, National Australia Bank and Commonwealth Bank.

But with the launch of a new digital platform for consumer banking in early September, it has made clear its commitment to grabbing a larger slice of the pie. And while it might be argued that those four competitors have a hefty head start on Macquarie in offering digital services to customers (ANZ, for example, launched Australia’s first Apple Pay service almost five months ago), Ben Perham, the company’s head of personal banking, argues that’s not the point.

Macquarie, he says, isn’t aiming to compete with those banks on digital. In fact, he insists, the company wants to stop thinking like a bank. Instead, it’s aiming to deliver a customer experience more akin to that offered by companies such as Spotify, Netflix or Facebook:

We looked at the technology being used by leading digital companies and we saw an opportunity to bring that level of personalized and intuitive customer experience to banking - an opportunity to build, from the ground up, a market-leading retail banking suite that doesn't look like other retail banking suites and that brings together details of a customer's transaction account, savings account, credit card and home loan all within the same easy-to-use digital experience.

So we see our competition as the last app a customer used on their phone.  Or, putting it another way, we want to be one of the apps they have on the home screen of their phone and that they enjoy using on a regular basis. That's the bar we've set ourselves.

Speaking the lingo

Two years ago, with that goal already in mind, Macquarie’s Banking and Financial Services (BFS) Group hired Luis Uguina from Spain’s BBVA as its chief digital officer, heading up the team working on new digital products. In developing the platform, Uguina says, they’ve worked on a number of features of which he’s particularly proud.

The first is natural language search, which  allows customers to type or speak a query into the platform and get an accurate response. For example, they might query, ‘How much did I spend on groceries last month’ and the system will return any transactions associated with groceries.

But these queries can also be more complicated: ‘How many grocery transactions did I make totalling around $100 in the Sydney suburb of Mosman last month?’ Uguina claims that, in terms of accuracy, nine out of ten queries are returned correctly and this continues to improves. The system, he boasts, also recognizes a good deal of Australian slang, so that customers can ask about spending at ‘Maccas’ (McDonalds) or ‘Woolies’ (Woolworths) or on items such as ‘sunnies’ (sunglasses) and ‘barbies’ (barbecues). 

The second is machine learning capabilities, which enable the bank to identify patterns around an individual customer’s spending behaviour and then predict their spending and cashflow based on that personal history. This is shown graphically to the customer via a calendar layout, allowing them to see at glance their likely financial commitments in the weeks and months ahead and giving them a sense of how much money it’s safe to spend.

At the heart of this platform is the Cassandra database and a number of associated tools that come as part of the Datastax Enterprise (DSE) platform, including Solr for search and Spark for analytics. Datastax’s DSE OpsCenter tools, meanwhile, provide the bank with monitoring and alerting capabilities. Cassandra provides a smart data storage layer that is fed with information from back-end systems within Macquarie through an open API platform and then serves customer requests with great speed, due largely to its in-memory capabilities.

The project to develop the new consumer banking platform at Macquarie began in January 2015. The first release of the beta version was rolled out internally to bank employees in June 2015 and then to selected ‘early-adopter’ customers in November 2015, says Uguina.

Today, says Perham, it’s running in three environments - Macquarie’s labs environment, a beta environment for employees and early-adopter customers, and in production - with a steady stream of new ideas flowing from the first environment down to the third, all aiming to help customers with personal budgeting and financial management, rather than simply listing transactions. Says Perham: 

Our challenge at this stage, quite frankly, isn’t finding new features and functions that would add value for customers - it’s prioritising them. We’re smaller than our competitors [in consumer banking] but we believe that this will help us move more quickly. In that respect, I reckon we’re thinking more like a fintech start-up than the established bank that we are.

A grey colored placeholder image