Australian mid-sized firms gain from AI, but skills lag

MYOB has published research on AI adoption among Australia's mid-sized businesses, pointing to productivity gains alongside persistent gaps in skills, governance and systems.

The study surveyed more than 500 Australian leaders and decision-makers in mid-sized companies. It examined progress towards what MYOB describes as business autonomy, where routine administrative work is handled with minimal effort.

It found that 81% of respondents reported a positive impact on productivity from AI. Among the most advanced businesses in the survey, that figure rose to 93%.

Businesses that had embedded AI into core processes were also more likely to report broader commercial benefits. The research found 77% of those organisations were seeing improvements not only in time savings, but also in output quality, revenue and profitability.

The results were assessed against five areas in MYOB's Business Autonomy Maturity Model: core processes, data landscape, AI strategy, AI governance and workforce capability. The findings suggest many firms have made more progress in digitising processes and improving data environments than in training staff, setting AI strategy and putting governance in place.

That uneven progress is reflected in several survey findings. While 77% of businesses said AI was embedded in core processes, only 66% had structured staff training in place.

Key constraints

Cybersecurity and data privacy concerns were among the most commonly cited barriers to wider AI adoption, named by 45% of respondents. Workforce skills and capacity for change followed at 37%, while 36% pointed to governance requirements.

The research also suggests older technology remains a brake on wider adoption. More than 80% of Australian mid-sized businesses said they were planning ERP upgrades over the next two years, indicating many see legacy systems as a constraint on further progress.

According to the findings, firms with modern cloud-based ERP systems and integrated data environments were more likely to report stronger commercial outcomes from AI. Those outcomes included improvements in revenue and profit margins, rather than gains limited to efficiency.

Paul Voges, Executive General Manager of Mid-Market at MYOB, said the data showed both progress and limits in how businesses were deploying AI.

"Australian mid-sized businesses are moving quickly when it comes to adopting AI and investing in modern systems and, importantly, they're already seeing the benefits flow through to productivity, revenue and profitability," Voges said.

"What the data suggests is that the biggest gains from AI are being realised by those businesses that have invested across all the foundational pillars. Where there are gaps, whether in workforce capability, governance or strategy, the returns from AI are more limited.

"A business with strong systems but low workforce readiness, or AI embedded without clear governance, may not be operating at full potential. The opportunity now is to strengthen those foundations to unlock more scalable outcomes."

Implementation divide

The findings point to a divide within the mid-market itself. Larger companies, especially those with more than 100 employees, were more likely to invest in structured training, integrated systems and governance frameworks.

That difference in preparation appears to be reflected in outcomes. More advanced organisations were significantly more likely to report positive productivity effects and commercial gains from AI than businesses at an earlier stage.

Oxford Economics Australia also contributed analysis to the research. Associate Director Alex Hooper said the results were notable at a time when productivity remains a central issue for many Australian businesses.

"Workforce capacity and skills are among the most commonly cited constraints to further progress," Hooper said.

"Businesses that reported strong training programs also reported stronger gains in productivity, revenue and profitability.

"Data quality and integration remain consistent barriers across businesses and are also among the most common areas of planned investment. Strengthening these foundations will be critical to unlocking more reliable, scalable and trusted AI-enabled automation."

Voges said the survey pointed to a widening gap between businesses that had moved beyond trial use of AI and those still building the systems needed to support it.

"Businesses that have safely embedded AI into their core systems are seeing more significant commercial impacts, while those relying on legacy systems tend to see more limited, efficiency-based gains," Voges said.

The broader picture from the survey is that AI adoption in Australia's mid-sized business sector has moved beyond experimentation in many cases, but the gains are not evenly distributed. Companies that have invested in data, systems, governance and workforce readiness appear better placed to turn AI from a narrow efficiency tool into one with a measurable effect on revenue and profit.

"This points to an AI implementation divide within the mid-sized business segment," Voges said.

"Businesses that have built a strong base across data, systems, workforce capability and governance are pulling ahead, because they are able to move beyond experimentation and embed AI into the way their business operates.

"The opportunity now is to help more Australian mid-sized businesses strengthen those foundations so they can scale AI effectively and unlock the full productivity potential and commercial benefits available."