Australian & NZ firms struggle with poor data amid AI ambitions

Australian and New Zealand organisations are under mounting pressure to realise business value from their data, according to Salesforce research. Data and analytics leaders across ANZ report a pressing need to overhaul data strategies as ambitions for artificial intelligence (AI) capabilities rise within businesses.

Data readiness gap

Salesforce research shows 83 per cent of Australian business leaders feel growing stakeholder demands to leverage data, but outdated and fragmented datasets remain the main challenge. Nearly nine in ten data and analytics leaders, 88 per cent in Australia and 89 per cent in New Zealand, say their organisations' data strategies require a complete rethink to enable effective AI deployment.

While 58 per cent of Australian and 64 per cent of New Zealand business leaders consider their organisations data-driven, a similar proportion-60 per cent in Australia-admit difficulty in using data to drive business priorities. Only 43 per cent of Australian leaders and 47 per cent of New Zealand leaders say they can generate timely insights from their data.

Concerns over drawing incorrect conclusions are widespread. In Australia, 48 per cent of data and analytics leaders say poor business context leads to misinterpretation, rising to 64 per cent in New Zealand. The most frequently cited barrier to becoming data-driven remains incomplete or low-quality data.

AI amplifies pressure

AI is now the top data priority for Australian technical leaders, having jumped from seventh place in 2023. This shift heightens scrutiny on the reliability of existing data foundations. Most data and analytics leaders, 76 per cent in Australia, feel a strong push for speedy AI implementation, but 41 per cent lack full confidence in the accuracy of AI outputs, which often rely on outdated or disconnected datasets. Among those with AI in production, 85 per cent in Australia and 95 per cent in New Zealand report experiencing inaccurate or misleading results.

Failures in data trustworthiness have also impacted resourcing. Across Australia, 56 per cent of those training or fine-tuning AI models have wasted resources due to flawed data, a figure that rises to 67 per cent in New Zealand. Strong data foundations are viewed as critical; 87 per cent of Australian data and analytics leaders agree this is the most important factor for successful AI adoption.

Siloed data challenge

Unified data is seen as key by over 90 per cent of data and analytics leaders in Australia, but application sprawl complicates the situation. The average company uses 897 applications, of which only 29 per cent are interconnected, fragmenting data and creating inaccessible silos. Australian leaders estimate 18 per cent of company data is currently siloed or otherwise unusable, while their New Zealand peers put this figure at 22 per cent.

Three-quarters of Australia's data and analytics leaders believe the most valuable business insights remain in this inaccessible data, a figure even higher in New Zealand at 84 per cent. This leads to limited AI capabilities, lack of customer insight, reduced personalisation and missed revenue opportunities for more than 80 per cent of leaders surveyed.

Technical responses

Status quo challenges are prompting changes in data access and management. Sixty per cent of Australian and 49 per cent of New Zealand organisations are turning to zero copy data integration, a method allowing simultaneous access to distributed data across multiple databases without needing to move or duplicate information.

The adoption of zero copy is showing benefits. In Australia, companies using the approach are 27 per cent more likely to deliver improved customer experiences and 173 per cent more likely to succeed with AI initiatives. In New Zealand, the advantages include 138 per cent greater connection of customer data sources and 207 per cent higher likelihood of succeeding on AI projects.

Technical leaders are also trialling natural language interfaces for accessing data. In Australia, 62 per cent find converting business questions into technical queries error-prone, a figure that rises to 85 per cent in New Zealand. Almost all business leaders-95 per cent in Australia and 97 per cent in New Zealand-agree performance would improve if they could interact with data in natural language.

Governance and security

Data governance and security are also in focus, with only 44 per cent of Australian data and analytics leaders implementing formal frameworks, compared to 49 per cent in New Zealand. Meanwhile, 94 per cent in both countries see the need for new approaches to governance as AI becomes more central to business operations.

"The lessons learned from earlier waves of AI adoption provide a blueprint for companies to become agentic enterprises where human employees and intelligent AI agents work together. Trusted, unified, and contextual data is the key that unlocks everything. For organisations ready to execute at scale, this is the moment to shore up data foundations to confidently scale AI to its full potential to deliver real value and ROI," said Michael Andrew, Chief Data Officer, Salesforce.