ANZx gets business to create its own next-best actions – Finance – Software

ANZx is ramping up the sophistication of a backend engine that oversees how the bank engages customers of its ANZ Plus digital-only offering with relevant communications.

Marketing technology lead Karan Mehta told last month’s PegaWorld Inspire conference in the US that the bank had deployed the Pega customer decision hub (PegaCDH) for communications from the outset.

“We wanted Pega to be the central brain that covers all the touchpoints of the customer, mobile being our primary channel of distribution,” Mehta said.

“Right from day one when the customer joins ANZ Plus, we wanted to send them a welcome pack, and take them on a journey with us, whether it’s engaging them with the new products we’re offering that helps them enhance their financial wellbeing, keeping them engaged and retained, [or] connecting them with [digital] coaches so they can get the help that they need.”

PegaCDH was set up in a way that would allow it to handle more types of communication and use cases over time, Mehta said.

Many of the triggers for putting a certain message in front of ANZ Plus users in the app are informed by “events” – customer interactions and behaviours – in real time.

“You can trigger an action on the back of an event, you can use an event to just capture data and not trigger or do any decisioning, and last but not least you can look for an absence of an event,” Mehta said.

That ensures, for example, that customers’ initial experiences with ANZ Plus are positive and go to plan – for example, that a welcome pack is received, and if not, that the omission or variation in expected communication pattern can be flagged.

Next-best action

Much of the event-driven communications take the form of ‘next-best actions’ – a common initiative seen in Australia’s banking sector to put the right message in front of a customer at the right time, and on the right communications channel.

Mehta said that PegaCDH was set up initially to put “highly personalised and targeted” next-best actions in front of customers in the ANZ Plus app.

The content of the communications generally relates to a “savings goal” that customers set at onboarding, and from their actions and behaviours after that.

Mehta shared that “47 percent of customers use at least one financial wellbeing next-best action”, suggesting the tailoring of the communication is finding its target.

Over time, the marketing technology team has worked to reduce the amount of time it takes from inception to implementation of the next-best action.

This has been achieved by consolidating steps – such as templating and testing – inside of Pega, instead of running them in other third-party tools.

Collectively, these steps have resulted in a “60 percent reduction in next-best action lifecycle” duration, Mehta said.

But this work is still centrally controlled by the martech team; ANZx’s next evolution is empowering business teams to create their own next-best actions, and to be able to access performance data after the fact on how those actions perform when placed in front of ANZ Plus users.

“If we look at all the use cases that we have, about 80 percent of them are really simple but are high volume,” Mehta said.

“It could be constant content changes that business expects, it could be just a minor iteration, or [they] want to do A-B testing, sending different content out to a different cohort of customers. 

“We want to empower business users to start taking advantage of the platform and build it themselves.”

Doing this requires some changes to allow the business functions to populate the customer data model – called the customer analytical record or CAR in Pega parlance – themselves, tapping into multiple data sources in the bank.

“We want them to push data into Pega CAR,” Mehta said.

“I think it’s going to be an industry-first pattern where we are moving away from centralised data dependency to a more federated environment to add data into CAR that would be used for decisioning.”

With the number of next-best actions being created and published also increasing, Mehta also said his team “want to push data out into the ecosystem, so our marketing and business users can leverage the data for measurement and optimisation” of the actions they create.

Data mesh contribution

Mehta noted this two-way exchange of data fitted into a broader push within ANZ Plus to adhere to “data mesh” principles around data management and exchange.

“At ANZ Plus we follow data mesh principles, so we want to move away from the philosophy of a centralised data lake where one team ends up becoming a bottleneck for everything.

“We want each domain to be responsible, to own their own data products or datasets, and publish it and make it available for the rest of the business. 

“So, in line with our principles of data mesh, as part of marketing technology, we wanted to push interaction histories of how customers are interacting with these actions back into the data ecosystem.”

Mehta said the data mesh structure essentially allowed business users and analysts to tap into data from across ANZ Plus and build their own views of it, specific to their own needs or purposes.