The first stage of robo-advice has been a “qualified failure” in Australia because the financial industry has used the technology to solve its own problems rather than the customers’, global actuarial and consulting firm Milliman has stated.

Milliman principal Craig McCulloch, whose company has powered a range of artificially intelligent advice tools for the industry, says the next phase of robo-advice should resemble Google Maps and be intuitive about where customers are placed on their financial journey and where they are likely to go.

“I think what we’ve done as an industry is misdiagnose the problem,” McCulloch says.

“Much of the first wave of robo-advice, globally, was asking how we automate existing processes and how we automate the way we deliver advice,” he says. “It’s been solving the industry’s problems, rather than solving the customers’ problems.

“It may not be the case that people need advice [the way it] has historically been given,” he says.

McCulloch gave his perspective during a panel discussion on robo-advice, with Zuper chief executive Jessica Ellerm and Willis Towers Watson head of digital solutions, Richard Body, at the 2018 Investment Magazine Post Retirement Conference, held in Sydney on March 20.

Milliman’s work with AMP over the last 18 months, building a digital advice platform, was based on the conversations advisers were going to have with customers and what they might need. It was an instructive approach for building robo-advice, he says.

“If we start with the customer, we’re going to get empathy much more quickly,” McCulloch says. “The example I often use is Google Maps. It’s a useful digital map but what makes Google Maps interesting is it tells me where I am and where I’m likely to get to.

“In terms of having access to spending patterns, we are understanding customers better than they understand themselves. Building a Google Maps for finance and genuinely attractive customised apps…that’s where we’ll see robo-advice become an unqualified success.”

Jessica Ellerm, chief executive of Millennial-focused fund Zuper Superannuation, says her company used a chatbot called Zena that interacted with members about their money and ambitions but did not provide robo-advice.

She says Millennials’ experience on social media of being targeted by advertisers based on the data they created from online interactions made them expect robo-advice to have a similar understanding of their spending habits and advice needs.

“You shouldn’t have to sit down in front of a platform and answer obvious questions. (The robo platform) should be able to ingest bits of information about you,” Ellerm said. However, McCulloch said the Facebook predictive model does not currently work for robo-advice because it is not 100 per cent accurate, causing problems with Statements of Advice and compliance issues.

The industry would be better to build robo-advice solutions first and then make them compliant, rather than beginning with compliance constraints, he argued.

“It’s a great aspiration but barriers to getting there are quite high,” McCulloch said. “There are many challenges before we get to that stage.”

Meanwhile Willis Towers Watson head of digital solutions in Asia and Australia, Richard Brody, says success for the next phase of robo-advice will come down to meeting the needs of customers of all ages and all stages of their financial journey.

Join the discussion