While interest in data analytics may have been eclipsed recently by the rising interest in artificial intelligence, its time in the sun is far from over. Analytics systems may lack the ability to learn on their own, but they still form the backbone of modern data-driven organisations and continue to be more reliable than their hallucinating AI counterparts.

Interest in AI is represented in the comparative spending between it and big data analytics. According to Statista, the global market for big data analytics will clock in at US$349.6 billion ($527.29 billion) this year, compared to projected spending on AI of US$184 billion ($277.52 billion). However, rapid growth in spending on AI will see it surpass that for big data analytics in 2028, with AI spending reaching US$826.8 billion ($1.26 trillion) in 2030 compared to US$655.5 billion ($988.67 billion) for big data analytics.

Building on strong foundations

The race to take advantage of AI has led many organisations to up their efforts to bringing their data assets together. Typically, this has led to a choice between centralising data into a data lake or lakehouse environment, or keeping data assets distributed while centralising their management and analytics functions.

It is this latter option that has been taken by ANZ, where chief technology officer Tim Hogarth said the bank has favoured a data mesh configuration. Business units take ownership and responsibility of their data assets, which they publish as data products for use across the organisation.

"A mesh is as much a business architecture as it is a technical one, and it is a response to how organisations have divisional centric structures," Hogarth said.

Hogarth said one of the benefits of the data mesh approach is the ability to more efficiently reuse data.

"You need to design systems that allow you to reuse data so that it can be packaged up and consumed by other divisions," Hogarth said.

"What I'm focusing on now is where we can find and mobilise our business teams to start to utilise AI.

"This technology has genuinely transformational capabilities, and it's going to require quite a different execution approach and quite a different mindset."

ANZ is currently experimenting with the application of AI to tasks such as identifying markers of hardship, helping it communicate faster with customers by interpreting conversations and texts, and assisting institutional banking teams be better informed about customers.

"There are much more significant use cases that still require a lot of business mobilisation, and that business mobilisation is a combination of both technical acumen and business ambition," Hogarth said.

Turning that ambition into outcomes however requires an ongoing investment into data literacy and sophistication.

"It's not a case of just simply rubbing AI on a problem, it's a case of identifying where you have a systemic advantage because of the nature of your data, and then working out where the new efficiencies can be gained by utilising machine learning and artificial intelligence," Hogarth said.

The organisational imperative

The need to better organise data assets was also a key consideration for Balamurugan P. M. when he joined the National Rugby League as CTO and general manager and for technology in April 2022 after his more recent role in banking.

According to Bala, the perception of data and analytics at the NRL is very different to that which he had experienced elsewhere.

"The way we look at data is very different here in sports, in terms of the perception of fans, communities, stakeholders and partners," Bala said.

"Our goal with technology is we want to be a strategically data-driven sports organisation going forward."

What Bala saw when he first joined was a scenario that is common across many organisations – data that was stored across multiple siloes – so he embarked on what has become a common set of actions within the analytics discipline. He describes the data strategy as a three-stage process based around the concepts of stabilise, accelerate, and grow.

"The first part of our mission was to connect these data points and provide more stable data capabilities to the business," Bala said.

"That has been our goal since 2023 - getting the right frameworks around data, doing data scoping and identifying sensitive data, and scrubbing it and masking it as well, and then putting in the right policies and procedures before exporting and migrating data."

Having completed the 'stablise' work in 2023, the 'accelerate' layer has involved building a data lake and connecting in the different data points.

Many of the tasks undertaken by Bala and his team represent the realisation of key trends in data analytics, from implementing effective governance strategies through to centralisation of data assets and improving the flow of data out to the parties that need it. But this is only the start of his ambitions.

"One of the ambitions we have is that by mid next year we are not going to have any static reporting going forward in the organisation, so nobody is going to create and build reports - we are building generative AI reporting," Bala said.

Bala said the goal is to create models that can bring in multiple data sets and use generative AI to answer queries quickly and effectively.

"For example, if NRL is planning to host a State of Origin game in Melbourne, and an executive wants to know the ticket sales forecast ... we will be taking all of these data points to start providing a forecast and report," Bala said.

"Normally that takes a few people to look at different reports and systems and come back in a week’s time. We want to give that reporting capability back to the users, so that is a goal we are working towards by June 2025."

A symbiotic relationship

Bala's goal of bringing greater use of AI into the analytics function is also paralleled by another of the key trends in analytics – using AI used to comb through vast troves of data to identify anomalies or patterns of interest.

This not only enables the analysis of much greater datasets than could be accomplished using human analysts, but it also enables those analysts to switch their investigations to outcomes and recommendations based on the insights that AI is deriving.

AI is also playing a significant role in making more data sets accessible for analysis, such as through converting unstructured data such as telephone calls into text.

So, while AI may be challenging data analytics for the attention of data professionals, it also holds promise for further extending the functionality of analytics processes.