top of page
  • Athina Mallis

Dr Alex Antic on key trends and challenges in data analytics

Dr Alex Antic, head of data science at the Australian National University discusses with Target Market podcast host Athina Mallis the key trends and challenges in data analytics and how organisations can strengthen and scale data capability to drive business success.

What is one of the common mistakes organisations make when attempting to scale?

It normally comes down to not giving enough consideration to the scalability element and not considering some of the key aspects. You really need to be investing in the three pillars which are people, tech and data. You need to be able to embed scalability into the organisation - by that I mean are they able to empower staff with data and tools, and be able to use that on a scalable level. That really comes down to a culture of innovation and change, to help promote that it comes down to senior support, having a CDO at that level in the C-Suite to steer in that support the initiative around data and around scalable data and analytics.

To do that effectively you need to be able to embed analytics into the decision-making process and this is once again where the CDO can play a really crucial role to ensure they're there to promote and advocate the use of analytics to inform decision making at the most senior levels. Getting the right people, tech and data in place, having the right culture to support that, often means the scalability element doesn't become a blocker in some organisations, but it becomes something that leads ultimately to your success.

There’s a lot of competition when it comes to choosing data analytics tools, what should organisations look for when they’re investing in the right solution?

It is often important for them to have guidance from trusted experts either internally or externally - especially for organisations who are trying to ramp up their analytics maturity. you need someone who can assess the needs of the organization, To be able to vet capabilities that are deemed possible by vendors either in the vendor space or open-source solution and tools, both infrastructure and software. To be able to ensure that those needs that happen at a strategic level actually met by the tools and infrastructure they are looking to implement.

Often I advocate strongly to have a suite of tools at your disposal having say open source and vendor-specific so that analysts and data scientists can find what is fit for purpose for the particular problem they have at hand. You don't want to standardise too much as projects can vary quite a bit but you want to be able to iteratively develop your capability and to be able to solve your business problems, you need to be able to be flexible in how you experiment and explore and implement the latest machine learning algorithms.

You need to make sure that the solutions you have, meet the needs of those technical experts that are working with them. Provide the right cutting edge technology and play to their skillset and demands, but also choose tools that attract some of the best people and retain those staff that want to be working with cool new technology, not outdated stuff.

Finally, it comes down to having tools to allow data democratisation so getting data and analytics into the hands of the broader organisation, not just the technical experts. That often means having scalable solutions, different types of processes and tools in place to allow people who may be at different levels of the hierarchy or technical proficiency to be able to look at the data, ask questions of it, and come up with some informative analysis.

There are a number of factors that you need to balance, but ultimately, it comes down to the people, purpose of that organisation and what helps get the data in the hands of as many people as possible.

You’ve no doubt seen a lot of data analytics companies market themselves in the APAC region, what are some examples of good marketing tactics data analytics organisations can implement to earn the trust of local analytics professionals?

One of the first ones is to have a good suite of use cases that you can use to show how you've leveraged analytics to really meet strategic goals whether for your own organisation or others that you're serving. Sometimes it can be done in an indirect way, such as sponsoring events such as hackathons, conferences, meetups, internships for students, community outreach especially within the STEM network - and showing you really care about the use of analytics beyond your own organisation and how the public actually perceives it and understands it.

It's also really important to have good trust in known people. For many analysts, the passion for them is to work with really strong people, people they can learn from people they can bounce ideas from, people they respect. Being able to highlight key people in your team who do promotion through that, through LinkedIn, through other media to get it out there to say, we really care about the type of people we have, we want to nurture them, we want others to come and work with these fantastic people.

Other aspects is offering helpful advice and assistance - that could be through blogs, podcasts, interviews and other events. You could be a bit more explicit in offering guidance to government bodies or NGOs or broader communities, assisting them with their own analytics challenges. The final one is around diversity, showing that you're a diverse and inclusive culture, you're aware of responsible AI in terms of ethical bias aspects and you're there to serve the public good in some way through the products and services you're offering beyond just your own financial goals that you have its broader than that.

Looking to the future, what are your top 3 predictions in data analytics for 2021?

Apart from continued growth and spending on analytics, which will continue post-COVID as more organisations realise the need to increase the analytics capabilities. I'm seeing increased use in unstructured data for strategic value, think along the lines of text data, social media data, video, audio across many different domains and fields, finance, government, marketing more broadly. Trying to tap into some of the unrealised value that's stored in some of this data.

Another one would be cloud-based technology. I think as organisations continue to scale it's becoming the norm these days, even in the government sector where predominantly hybrid cloud is being a solution given some security constraints that they have but that's changing. I've seen that change over time.

Cloud overall will become much more common as a scalable and viable option for organisations as they grow their capability and data holdings and how they actually explore that data and derive value from it.

The third one would be around privacy, I think it's going to become more of an issue where we have this constant tug of war between social license, versus regulation legislation, versus public good.

As legislation can't keep up with technical advances I think we need to use technology to work within the right boundariues. So things like confidential computing privacy-enhancing technology, these techniques allow us to tap into the value of underlying data without disclosing any information. I think they'll become a lot more prominent in use across many different organisations.

I'm seeing currently an increased use across both public and private sectors, which have been really exciting. I think there are going to be great findings and huge value to add to many organisations in the next year or two.

Final thoughts - what’s a commonly held belief in data analytics that you passionately disagree with?

The one thing is people think it's about the tools and the technology, to me it's about the people. Data science and analytics to me is about change, transformational change within an organisation to enable data-driven, evidence-based decision making culture.

The tools and technology, they're just a tool - a means to an end. I think you need to get the people part right, you need to have the right people in the right roles, they need to actually formulate the questions that you're trying to solve, they're the ones that lead to strategic value for an organisation, they are the ones that support the analytics and they are the ones that ultimately make the decisions that matter.

Getting the right people, having the right culture to support them, can make or break for most organisations. I think it's absolutely paramount to get that part right. Worry about the people first, then the technology later. That's I think is the most important part.

What should everyone who is involved with data analytics start doing?

With 'start doing' is to give a lot more consideration, a lot more thought to the notion of responsible AI, to think more deeply about how is what you're developing from an analytics perspective - how can that be used for both good and evil? In the broad sense, how can that really impact people's lives and what's the responsibility to ensure it is used appropriately?

Thinking about bias and fairness, thinking about how to leverage the capabilities to achieve public good and creating the right ecosystem and processes in place to support that and a part of that is diversity within your organisation. Diversity of thought and diversity of people, culture and really having a cross-disciplinary team that can really add value to this is absolutely paramount.

On the flipside what should everyone involved with data analytics stop doing?

'Stop doing' is something I see a lot at senior levels is to stop buying into all the hype and misconception around AI that it's going to solve all our problems and robots are going to take over. That completely misses the point, it takes us down the wrong path. It's really about how we can work together with machine learning and AI to help solve some complex problems, to help deal with problems at scale.

A big part of that is focusing on clearly defining the business problem we have, making sure it's a viable problem to solve via the analytics paradigm. Then to start simple and building complexity as we need it. These techniques are there to solve business problems, people are the one that identify the problems and that ultimately makes the decision and people are the ones that define what is fair, what is bias and what is morally correct. I think it's important to focus on why we are here and what we are doing is to deliver value through services or products to customers and people in general and to forget the people part.

Tune into the full episode here.

Target Market, a podcast series by AZK Media, where the world’s most premium thought leaders across technology, marketing and data come together to share their insights. Hosted by Athina Mallis.


bottom of page