'Disrupt or die': Why customer experience is the new battleground at Macquarie Bank

The bank is rolling out mobile apps and chatbots to keep customers engaged, explains CDO Luis Uguina

"You either disrupt or you'll be disrupted," said Luis Uguina, CDO of Australia's Macquarie Bank. "This goes for every sector." As an example he mentioned the retail sector which has been turned upside down by newcomers such Amazon.

"We can all learn some lessons from what's happened to retail," he went on. "The customer is in control now, and there is no more loyalty. You have to compete on the basis of customer experience."

Before embarking on a technological transformation process to support Macquarie's own attempts at disruption, Uguina and his colleagues spent some time asking the its retail banking customers for their views.

"We're the new kid on the block. What should we be doing? What should we build next? We told them to think big, to tell us what they expected to see".

In summary, customers' vision of the future was a much more personal service, mediated by new channels of communication.

The personalisation paradox

In retail banking, as with other sectors, personalised service is seen as a key market differentiator. If they don't already, customers will soon expect to be able to communicate with their bank via a mobile device in more or less the same way as they would a person, asking questions and receiving answers in their own language rather than in the lexicon of the business. Uguina gave some simple examples of such natural language queries.

"Show me how much I spent on my last trip to London. Find me the receipt for the Xbox I bought two years ago."

Paradoxically, of course, this "personal touch" must be delivered without any human intervention - given the speed of response required and the scale of the coverage that would be impossible. Instead, we enter the realm of machine-driven "mass tailoring", with vast quantities of data processed for context and acted upon in real-time.

"Big data is necessary but not sufficient," Uguina said. "At the same time real-time is necessary but not sufficient; you need both things".

The infrastructure required to support real-time personalised interactions is very different from the mainframes that have traditionally underpinned the bank's operations. But rather than rip and replace the legacy infrastructure, which still handles batch tasks perfectly proficiently, the bank decided to build an extra data layer on top.

"We created a smart storage platform which supports our machine learning algorithms and real-time processes," Uguina explained. "When a customer uses a credit card then the authorisation gets sent to this layer. Then we have all these processes analysing the data and triggering new alerts for the customer according to what they want from the bank."

A moving target

Machine learning and natural language processing - which provide the basis of Macquarie's disruptive potential - are both areas in a state of flux, with big advances posted every few months. This can make it challenging to plan ahead.

"We deliberately chose tech that would be future compatible, that's constantly being updated," said Rajai Rai, head of digital engineering at the bank.

"So it's built on DataStax Enterprise, with Cassandra, Spark and Solr for full-text search. Plus we use microservices architecture and APIs that are open to new technological developments. All these technologies lend themselves to agile."

The first release, a minimum viable product (MVP) of a mobile app, was rolled out after six months' development time, with further iterations informed by technical and customer feedback. Ten months later all the data on the bank's legacy mainframes has been replicated to the new smart data platform, allowing them to be phased out, should that decision be made.

The new layer now features an API platform and machine learning algorithms and supports a mobile e-banking app, a chatbot and desktop applications. Additional features are being added all the time as the algorithms learn more and as voice processing technology by the likes of Google, Apple and Amazon advances.

"This is the new world. Everyone's got a mobile and they're connecting 24/7 so it always on," said Rai. "Even when it's being upgraded it's on. There's no such thing as a maintenance window anymore."

The smart data layer gives the banks the flexibility to be more responsive to the changing expectations of customers, Uguina said.

"Customer experience is no longer a playground, and hope is not strategy," he concluded. "Nothing will happen unless you improve customer experience, and if you don't change you will die".