Large enterprises often struggle to replicate the innovation of their industry's disruptors, but the Macquarie Group has launched a new retail banking service that proves it can be done. Chief Digital Officer of the Australian financial services company Luis Uguina decided to develop the new digital platform because he believed the big bank could compete with the customer experience provided by the digital natives.
"The idea was, as we are launching a new bank, what we would love is to deliver exactly the same customer experience that you receive when you are dealing with pure digital companies," he told CIO UK over breakfast in the UK's tallest building the Shard.
"Usually a big disabler is the technology, legacy systems, old technology, a lot of spaghetti in the back end systems. Because at the end it's the result of 50 or 60 years of putting new systems and trying to update the system.
"We are using exactly the same technology as companies like Apple or Netflix are using. Why are we not able to deliver exactly the same customer experience? That was the idea, and we have been working over the last 16-18 months to deliver the retail banking service."
To deliver this level of user experience, Macquarie its approached customers to find out exactly what they wanted from their bank, beyond the standard banking services in a secure environment. The answer was a new way of searching for information.
"What you're asking the customer is to think of the nirvana, something that you would love working with," Uguina recalls. "What we received was that when we work with a bank, we would love to be able to speak to the bank in the same way that we speak as humans."
The recent advances in natural language processing made the technology an obvious choice of interface to explore and fulfil this desire, but the information they required was as important as the method of getting it. Macquarie customers wanted a simple way to find out details on specific expenditures, such as the cost of the breakfast in a London restaurant like the one Uguina was eating, and they wanted it delivered in a personalised manner.
"What they told us was, we would love to have a bank that is aligned and is tailor-made for me," he says.
"The bank should always understand my behaviour, and should be able to deliver an experience that is aligned with my expectations, because usually what happens is that the banking industry delivers a one-size-fits-all. The bank should be able to meet every single customer's expectation, across all the different products and across the lifecycle of the customer.
"With those requirements, we've tried designing the system from scratch. What we have delivered is what we think is the best customer experience that you can get in a bank worldwide, and we have implemented everything that the customer wants."
The resulting platform combines all the bank's products in the same single space to support a seamless user experience. All the different channels offer natural language search technology, so customers can request minute level of details as if asking another human being. A businessperson travelling abroad, for example, could ask what they spent on a single transaction or during a specific time period, and at the end of the trip request a report compiling all the expenses he or she had accrued.
"The most important thing is that we deliver a single bank view to the customer. Regardless of if you have six products or 60, or if you have a home loan or a credit card or debit card, you see the bank as one single bank," Uguina explains.
"We are using angular technology, to be able to deliver a brilliant customer experience. We are using native applications for IOS and Android, and then all of the systems can use the website, so it's a fully responsive website."
Macquarie is constantly fine-tuning the system, by monitoring the different channels to analyse the user behaviour and by responding to the growing number of customer requests.
The natural language search use showed that the workings of the human brain are more complex than the equations on which most systems are built. To calculate spending over the festive period, for example, the customer would ask how much was spent during Christmas, rather requesting transactions between specific dates.
The taxonomy of the system was updated to reflect this. It now understands that Christmas means sometime between 21 December and 7 January, and gives customers the chance to verify that these dates are appropriate.
"This is what the people are really doing, and this is the way the human brain works," says Uguina. "What we are doing is we are monitoring all the different things that we are not able to deliver to the expectation of the customer, and we are evolving the system every single week to be able to deliver those capabilities."
The number of requests has grown as the system has evolved. Some customers wanted to attach documents such as receipts, and to keep all of their records in the same single place. Others would need an invoice to claim for a warranty and could remember the place where the product was bought but not the date at which they bought it, so they wanted to search for the shop where the product was bought.
Drawing on this growing quantity feedback due an improved customer relationship lets the bank adapt its service to the way its users think.
"When the customer knows that the bank is really trying to deliver or to solve the customer expectations, their relationship with the bank changes dramatically, and what happens is the customer starts to send you millions of requests," says Uguina.
"The idea is that we need to evolve with the customer and we need to evolve the customer experience every single day. There is nothing worse than asking the customer for feedback and then doing nothing. So what we have done is the whole bank has changed the approach to be able to deliver small chunks of functionality.
"We need to be able to merge our business plans with what's important for the people, because at the end our mission is to make life easier for our customers so the only way that we can make life easier for them is being able to deliver exactly what they need and what they want."
Uguina wants to fully understand the evolving needs of Macquarie customers as they change, and constantly and quickly react to their requirements. He calls the approach "doing business at the speed of now".
Macquarie operates an aggressive Agile methodology, with the whole company committed to tri-monthly programme increments and weekly updates to the existing systems. The bank monitors the success of each individual process and uses the data rather than ideas to improve the whole system.
"We are using DevOps and we are using continuous delivery and continuous deployment technologies," explains Uguina. "We are moving all the different software and products and services into open sieve, containers technology to be able to deploy faster and with better control of what we're doing."
Consumer tech inspirations
Macquarie draws on the specific functionalities of a various consumer platforms as inspiration for the features it develops for its customers.
"In terms o freal-time analytics and understanding the patterns and behaviour of the customer, Netflix probably is the best one," says Uguina. "In the way you deliver the customer experience and as an aspirational brand, Apple probably. In the way you manage the information in a completely different way and you are able to understand the customer and moving massive amount of information, Google.
"You also have many other startups in the fintech space that are doing things really well. The biggest challenge that every single fintech company has right now is they don't own the source of the information, so if you are a fintech company you still need a bank that is going to provide all the capabilities.
"The beauty for us is we are able to use the same technology that they are using and deliver exactly the same experience that they are, and we also own the platform so we know we are really good with banking."
Developing the software internally means Macquarie has to attract and build high-calibre tech talent to support a long-term strategy, although the complexity of natural language search led Uquina to seek an external supplier. The foundation of the system is an existing library from Stanford University.
"On top of those natural language processing libraries, we are creating the taxonomies and the domains for the financial industry," he says. "That's our idea. We will take the best technology out there in both open source and not open source, but on top of this technology we will create all the different products.
"As our mission is to make life easier for our customers, what we need to do is we need to integrate with the customers lives in the simplest way."
This isn't always the case with contemporary banking practices. The current method of paying for a coffee, for example, involves getting out your card, inserting it into a Point of Sales Terminal (POST), entering your PIN and then collecting your receipt.
Uguina wants the payment to be made and the receipt delivered in real-time using just a mobile phone, and when the system understands that the customer buys the same coffee in the same place every day, they customer should be able to automate the transaction process.
"It's currently a collection of different steps completely disconnected. You are using the financial product every single day, three or four times per day, but the whole experience is disconnected," he explains.
"The system should be able to understand my behaviours and my patterns, and should be able to create this seamless experience and frictionless experience that we don't have right now."
Changing company culture
Macquarie has long placed innovation at the centre of its operations, and a highly-supportive boardroom makes it easier for Uguina to implement new technology solutions. That doesn't mean it's always straightforward when the approach is to challenge every established process.
"What really creates a lot of discomfort is when people are not aware of the big picture, and the big picture is we are going to create a bank that is completely integrated with the customer's life. We need to create a bank that is working in real time," says Uguina.
"I want to be able to be a customer of the bank in five minutes, and be able to download the credit card to my phone in two more minutes, so in five minutes the whole experience has to happen. This is what I want. I want to create a bank where everything is happening now.
"If I want to open a new account it's not going to take 24 hours or 10 minutes, it's going to take 10 seconds. And if I want to search for the amount of money that I have spent in different categories in the last two years, I want to receive the information in less than one second.
"When you deliver this vision obviously you still create friction, but then people see why it's important. You are not breaking the bank, but you are challenging every single decision that has been working for the last 50 years."
Later this year Macquarie will implement another vocal component, with the launch of voice biometrics for customer security. Uguina is also keen to use graph technology from DataStax at Macquarie to incorporate external information into the bank's own deep dataset, to understand how current actions will impact future behaviours, and use the information to provide highly customised alerts.
If a customer sent their child an annual sum of money for their birthday, the bank should soon be able to remind the customer of the birthday in advance. Users would be able to decrease the AI involvement if the level of data it required made them uncomfortable.
"If you are happy with this approach where you receive a much better customer experience because you allow the company to understand you better, we are going to be the bank for you," says Uguina. "We are going to create exactly the same seamless experience as a Netflix, but on the financial side of the world."