Strategic use of data analytics has helped Macquarie Bank reduce its cost to serve customers by almost 50 per cent, the company’s chief data officer (CDO) revealed as the bank increasingly leans on machine learning and analytics to define its digital future.

That future has evolved over the course of a five-year digital transformation that has expanded across the bank’s Banking and Financial Services (BFS) group since it launched a “personalised and intuitive” digital-banking offering back in 2016.

Five years later, the industry’s rapid digital transformation has seen it developing a fully digital ecosystem, including features such as API-based extensibility, that has been embraced at every level of the organisation.

New products aside, the bank’s back-end cultural transformation has made the most enduring change in “empowering users”, chief data officer (CDO) Ashwin Sinha explained at this month’s Tableau Live conference.

“It was very important for us to have a strong data foundation that could support scalability, agility and robustness as we went on scaling our business,” Sinha explained. “We wanted to make sure that we were making good use of data to improve client service and client experience by personalising the interactions we are having with them.”

A former KPMG management consultant, Sinha joined Macquarie Bank nearly 2 years ago – inheriting a nascent job title that has, in many organisations, suffered from being relatively ill-defined and intermittently supported.

One recent survey found that almost half of APAC public-sector CDOs still don’t fully understand their role, with just 44 per cent saying their agencies were relying on data insights when making mission-critical decisions.

Sinha, however, has had no such issues thanks to a highly engaged executive that had understood the importance of data-driven culture long before he arrived.

Board meetings, for example, incorporate Tableau-powered live data dashboards that highlight a broad range of key performance indicators, giving executives the most current data possible to support their decision-making.

Given the ever-expanding number of regulations and risk-management requirements faced by financial-services companies, easy access to crucial data has proven transformative within Macquarie – driving the development of a relatively mature data culture that has, a “really proud” Sinha said, provided unprecedented visibility of “our performance, key improvement opportunities, and drivers to understand what are the steps and actions we could be taking.”

“We are making better use of data than ever before,” he said, “and reduced our cost to serve, in terms of cost of the platform, by almost 50 per cent with better use of cloud technologies and machine learning.”

Building a data culture

Macquarie invested early in structured analytics and data-driven decision making, with an established data culture that has embraced analytics for areas like sales and marketing, client service, and fraud prevention – ensuring that staff can, as Gartner recently put it, “speak data”.

The bank’s mature data culture has become a sort of breeding ground for skilled data executives, with UK bank Barclays recently reaching out as far as Australia to recruit former Macquarie Bank chief analytics officer Sheetal Patole as that company’s CDO.

CDOs “must become change agents”, Gartner advises, “focused on the transformational impacts of data-driven culture and data literacy.”

Four key factors, Sinha said, had shaped Macquarie Bank’s success in data-driven transformation.

These included a recognition that fast-evolving data tools and best practices “are going to change in a couple of years’ time – so it’s important that when you think about the data architecture, it is flexible enough to accommodate and adapt to these changes.”

Limited availability of “practical expertise” in areas like data management, data engineering, data science and data governance, he added, demands consideration of how companies will acquire, nurture and support data talent over the long term.

Managing user expectations was also critical, he said, including accepting that human factors mean compromise is inevitable.

“When you go on this transformation journey, you’re trying to have a perfect state,” Sinha explained, “but at the same time, user demands will require a level of tactical solutions – and these demands are genuine.”

Finally, he advised, it is important not to get carried away with data – which can easily become hard to manage if it is collected without a clear purpose.

“Question things that are not needed,” he said, “and question the purpose of every data set and data artifacts – and what actions these artifacts are going to support. Otherwise, you could create a lot of unnecessary assets.”