Accenture's Catherine Garner: "I am an advocate of doing things in bite-sized chunks because it's ok to fail - as long as it hasn't taken you 3 years and cost $50 million."
Governments are failing to make the best use of emerging technologies because they are working to unrealistic time frames that don't address a period of co-existence between new and legacy technology and processes.
This is the view of Catherine Garner, who leads Accenture Australia's Health and Public sector business. Garner told CIO Australia there are three barriers to the successful rollout of new technologies in government: an absence of in-house skills; lack of leadership; and legacy systems, processes and thinking.
Public trust in government services is continuing to waver as projects run over cost and behind schedule. Earlier this year, it was revealed that the Department of Human Services' $104 million child support system was more than 12 months late and in disarray. Acting Commonwealth Ombudsman Richard Glenn last month highlighted poor project planning, system testing and risk management in the implementation of Centrelink's online automated debt system.
"The three challenges that we see [for governments ] are around skills, leadership and legacy. To navigate today and in three months' time where you want to be, 'where's the skills, where's the leadership, where's the understanding of the legacy to get you there?
"It's getting the balance between those and not just swallowing the 'kool aid' from someone, [who says], 'don't worry, we'll set up a new CRM and you can migrate all the data and it's going to be straight through processing and it's only going to take six months."
Accenture's recent 'Emerging Technologies in Public Sector' research found that only 18 per cent of Australian government agencies reported having machine learning skills, compared to 62 per cent in Germany, 53 per cent in the UK, and 52 per cent in the US.
Garner said that Australia governments - and private enterprises - have challenges with experience at all levels. They must develop in-house skills and recruit specialists in areas such as machine learning, artificial intelligence, and biometrics - and make these roles attractive to internal staff.
"There are also lots of people talking about leadership and making one line statements about it, but there are not many leaders who actually understand the critical parts of [technology transformations] to deliver something for citizens and employees. We've got to address the leadership tension around this," said Garner.
Sign up for Computerworld eNewsletters.