The idea

Asana's State of AI at Work: Australia 2025 surveyed 2,010 Australian knowledge workers and found that 46% are using AI tools weekly. Only 14% of organisations have scaled AI across operations. Deloitte Australia puts the transformation rate at 12%. Deloitte Access Economics puts it at 5% of SMEs fully AI-enabled. These three studies measure different things: scaled deployment, transformation, full enablement. The convergence is the point. Adoption is everywhere. Returns are rare.

Why it matters

The competitive question has changed. It is no longer whether your business uses AI. It is whether you are measuring inputs or outputs.

Most businesses are measuring inputs. Tool adoption, usage frequency, time saved per task, percentage of staff using ChatGPT weekly. Inputs are easy to count and create the appearance of progress. They do not tell you whether the business is making more money. MYOB's October 2025 mid-market survey found 82% of Australian businesses using AI report positive impact, while 46% do not measure impact at all. The two numbers describe the same problem.

The pattern shows up at every business size. In May 2026 the President and COO of Uber, Andrew Macdonald, said this publicly. Ninety-five percent of Uber engineers use AI tools monthly. Seventy percent of committed code is AI-generated. The company burned through its entire 2026 AI budget in four months. Macdonald said he cannot draw a line between that activity and the useful features the company ships. That link, in his words, is not there yet. In Australian mid-market data, Snowflake found 81% of firms struggled to demonstrate the value of their AI investments. Different businesses, same problem.

Using AI more is not the same as getting more from AI. The first creates a bigger invoice. The second creates a different business.

The 86% have ChatGPT subscriptions, automated email drafts and faster meeting notes. The 14% have rebuilt parts of the operation around AI and can point to specific commercial outcomes: cost per file down, revenue per FTE up, pre-approval to settlement faster, shortlist cycles shorter, time to quote compressed. Same tools, different posture, very different business results.

The Alvo take

Most Australian businesses are mistaking tool use for transformation. A team using ChatGPT to write emails faster is a team writing emails faster. The customer onboarding sequence, the quoting workflow, the placement pipeline, the broker file, the listing process, none of it has changed. AI is layered on top of operations designed before AI existed. The invoice goes up. The business looks the same.

Stop measuring inputs. Pick three operational processes where the commercial outcome is measurable, cost per file, days to shortlist, time to quote, conversion to settlement, gross margin per job, and redesign them around AI rather than accelerating them with it. The result is not a faster version of the old business. It is a different cost structure, a different capacity ceiling, a different competitive position. That gap widens every quarter the rest of the market spends bolting tools onto unchanged workflows.

In three years the businesses making money from AI will not be the ones using it the most. They will be the ones who redesigned the work around the outcomes that matter.

Sources: Asana State of AI at Work: Australia 2025, surveying 2,010 Australian knowledge workers. Deloitte Australia (12% transformation rate, 2025). Deloitte Access Economics (5% of SMEs fully AI-enabled, 2025). MYOB Mid-Market Survey, October 2025 (506 Australian businesses). Snowflake ecosystem analysis (81% of Australian firms struggle to demonstrate AI value, 2026). ServiceNow AI Maturity Index 2025 (only 10% of Australian enterprises feel ready to reorganise their business with AI). Andrew Macdonald, Uber President and COO, Rapid Response podcast, May 2026.