Macquarie Bank uses transformation to simplify data environment

Macquarie Bank has undergone a significant data transformation that saw it consolidate onto a cloud-based platform, halving its operating costs and driving data deeper into its processes.

Chief data officer Ashwin Sinha told the Gartner Data & Analytics Summit that the bank set out to drastically simplify its data environment, with a focus on three key areas.

"How do we make our data landscape simple, and thereby reduce the cost, as well as reduce the risk in terms of running that data landscape?" Sinha said.

"The second one was around how we look at improving continuously the client and employee experience with better use of data.

"And the third one was around how do get even better with respect to use of data for risk management."

On simplification, Sinha said the bank consolidated "eight different data stores to a single data platform" that is hosted in the cloud.

The actual platform comprises a mix of "cloud services, commercial software and an open source stack", though it is architected to be flexible.

"The technology and the methods in the space are changing very rapidly, so it is quite important that you do not get very stuck with a certain way of doing things and you are able to evolve that," Sinha said.

The bank took the opportunity to identify redundant or otherwise unused reports, dashboards, extract-transform-load (ETL) tools, database schemas and the like, and to remove them.

"With that, we significantly reduced the data landscape which we had to run, manage and support," he said.

"That helped us reduce a cost of running the data platform by almost 50 percent.

"It has also improved the overall resilience and robustness of the data platform, and the risk management around it."

Outside of simplification, the transformation also sought to make improvements to customer experience (CX) and employee experience (EX).

On CX, the enhanced platform has a role in addressing and resolving customer complaints, understanding net promoter scores across support channels, and aiding fraud prevention.

"With respect to employee experience, it was about better access to data to drive lot of decisions," Sinha said.

From a risk management perspective, Sinha said that the bank was "already good in this area" but saw opportunities to use data "to get even better".

Efforts were focused on fully understanding how specific critical data elements related to the broader process from a regulatory and compliance perspective.

Two-speed transformation

Sinha said that the bank took "an almost two-speed approach" to the data transformation, allowing it to pursue longer-term goals while continuing to service existing business-as-usual demand for data.

"As you go through a data transformation journey you have got this perfect end state in mind - one single platform supporting all user needs and client needs is the sort of ideal goal," he said.

"But it is quite important that you continue to satisfy the user demand as you are going through that transformation.

"In our case we had almost a two-speed approach where the analytics and the data science team continuously provided the organisation, and different business lines and different functions within the organisation, with the data required to run the business, make the required decisions and the changes.

"In parallel, we were working on transforming the data platform and the data landscape to make it simpler, much more efficient and a much more robust and reliable platform."

Key learnings

Sinha said that a key part of Macquarie Bank's success in the data transformation was setting expectations for what could be achieved and putting in place metrics to measure progress.

"There is a lot of perceived benefit that people think [exists in a] data transformation, but what was very important for me was to make those benefits very tangible early on," he said.

"[We wanted to] understand how much cost we were going to really reduce, what aspects of client and employee experience we were going to improve and how we were going to measure that, and what aspects of data governance we were going to improve and how we were going to continue measuring that.

"So those were some of the key areas which we focused on."

Sinha said that an agile approach to program delivery was also important so that changing priorities or needs could be quickly accommodated 

"It's important that you operate in an agile delivery framework, because it makes it a lot easier for you to shift according to those priorities as you go through that journey," Sinha added.