Funds may suffer “paralysis by analysis” with too much data

While superannuation funds are a treasure trove of data on members and engagement, issues with both access and analysis capabilities prevent them from leveraging that information to improve member outcomes.

First State Super’s Amanda Ralph, AustralianSuper’s Peter Treseder and HESTA’s Georgie Obst all agreed at the Conference of Major Superannuation Funds today that they all had plenty of data, but that this led to what Treseder called a “paralysis of analysis”.

“A lack of data isn’t the issue … in fact at times there’s too much data … but it’s about being smart and being targeted about what data will give you an actionable insight,” Ralph said.

Obst said that this was particularly difficult for larger funds, which had greater resources to collect and analyse data than their smaller counterparts but also needed to collect so much more to get it to a point where it was statistically significant.

VIEW ALL

View all

Some funds also faced problems in data collection. “Every single fund”, for example, that Laneway Analytics chief executive, Grant Callaghan, had talked to, complained that they couldn’t get data from their administrator.

Ralph said the frequency of data collection was another issue: “once a year is difficult to be getting that Australian Prudential Regulation Authority (APRA) data, and that’s something where we probably could do better”.

The panellists also agreed that superannuation funds needed to work on utilising data in a way that would create actionable and accurate insights once they managed to gather the information.

“Data integration is a real problem in terms of bringing all that data into an integrated layout where the sources and assumptions of that data are known,” Ralph said. “And part of that process when building out your analytics capability … [needs to be] spending some time to get an aligned and agreed definition of data across the business.”

Obst emphasised that funds needed to be open to testing data to ensure its insights were reliable, at the same time as turning their minds to traps in testing methods.

“You’re never going to know if data insights are completely right, but you need to go out there and test it. And within that you need a culture where failure is okay … because data isn’t going to give us the whole answer,” she said.