Currently, Walter leads the business of KINSHIP digital www.kinshipdigital.com in Victoria, with an emphasis on customer success management. They help clients understand and implement the changing world of enterprise social media management platforms as they embrace social care, marketing, social paid advertising, PR and social data research (what is becoming known as the “Experience Cloud”). Social data analytics plays a vital role in the efficient use of these platforms and that’s a particular interest of Walters.
Walter’s business experience as a CIO, an internal auditor, in corporate planning, a VP Internal Business, and many roles in strategy, when combined with his technical background in mathematical statistics and computing science, provides him with a particular strength in be able to assess the business value of new technology.
1. How did you find yourself working in Data?
My very first job was for 3 years in the quaintly named “Computational Laboratory” of the Statistics Dept at UNSW. I moved on to spend 8 years as a Consultant in Computational Statistics at the University of New England. I enjoyed applying statistics and helping the academic researchers understand – before it was too late – the principles and limitations of experimental design. They were all better qualified than me but not in statistics and computational statistics.
Although I worked extensively with SPSS, I also did a significant amount of coding working in collaboration with the Professor of Computing Science who was a statistician. Later during that period, I was also appointed to the role of Chief Systems Programmer at the University Computer Centre.
I left that work because I felt that “data science”, at that time, was too far from the action. I wanted to be at the place where statistics and computing science created value and made money for business. I went from over 10 years working in universities to joining BHP as an internal auditor. That role provided me a platform to leverage myself into a long career in IT including as CIO of Asia Pacific Minerals Division and VP Internal Business for a for-profit unit for which I had devised a business strategy while doing a one year stint in Corporate Planning.
More recently I have been working with enterprise social media platforms and social data analytics. This is a fascinating field as it is evolving rapidly and “marketing technology” is complex and requires a sound business perspective to sort out the wheat from the chaff.
2. Who has been the biggest inspiration (mentor) in your career and why (within data)
Perhaps surprisingly because it was a long time ago, Associate Professor Jim Douglas who first employed me had a profound effect on me being accountable for the results of my work. I had no way out but to produce results for him which satisfied his detailed scrutiny. He also encouraged me to explore for myself whenever I had completed what he required of me. His encouragement instilled in me what became a lifelong interest in statistics and computing science.
3. What personal development do you do to keep yourself sharp?
Since the time I ceased directly applying statistics – 15 years before R was released – the whole data analytics world has completely changed. I count myself lucky that I had an excellent academic education in both statistics and computing science and had key concepts embedded in my mind which have served me well up to now, and will do so into the future. I have no trouble placing new technologies and methods into my mental models.
I am focusing on achieving practical familiarity with the business application end of the data science spectrum, not the data engineering end. I’m quite fluent in R and its application to statistical and machine learning. That allows me to communicate with modern data scientists. However, my real focus is on where I get most excited and that is on the bridge between the data scientists and embedding their solutions how a business operates. That’s what I study – use cases, reducing resistance, gaining trust, and communicating value.
4. Do you find that organisations are changing in the way that they view data and how it can be used for strategy?
The leading data-driven organisations such as Facebook and Amazon and Uber are truly inspiring businesses. While not every organisation will be a data-driven business I think that a lack of holistic and systems thinking about the power of technology – including data science – is leaving a lot of significant Australian companies vulnerable to disruption.
That’s a good thing in my view, because we need disruption and we certainly need more competition. As to how it can be used for strategy, in my experience that relates more to an organisation’s culture than any form of technology. It refers to creating the bridge of trust between technology and those at the coalface doing the work. That’s why I focus on bridging that interface and reducing resistance, building trust and enabling the integration of the science into how things are done day-to-day.
5. What can be done to ensure Australia is a market leader in data analytics?
Great question. In my experience the best way to encourage more investment in data science technology and skills in Australia is for it to be seen to have created outstanding business value and to have crushed those too slow to embrace it. Envy and fear create action and momentum.
6. How do you see technology in data progressing over the next 5 to 10 years?
Thinking about ten years out hurts my head so that I won’t go there. Over the next five years, we’ll see the rise of the “data science stack” by which I mean a cluster of enterprise-grade data science software. This will be offered with the ability to fully integrate into the other key business clusters such as the marketing, customer care, e-commerce, CRM, ERP, content management clusters.
For example, think of it sitting alongside clusters that are currently being labelled as Customer Experience clouds. The big shift from today is that the key entities and their metadata of labels and tags and trackers and reporting identifiers will be visible to the data science enterprise.
I may be completely wrong about this following opinion, but I think that deep learning will become an integrated part of the enterprise data science cluster before machine learning and statistical learning methods. This is because of the latter – despite the rise of the Citizen Data Scientist – still, require an understanding of the underlying assumptions as opposed to autonomous learning.
And that completes the loop right back to my earliest days when I patiently explained the statistical experimental design to PhD candidates who came to ask me how to prove their hypothesis from their horribly confounded data which they had been collecting for the last 4 years.
That process of explaining from a perspective of analytics and business is the brightest star on today’s data science horizon.