A new Australian report found 89.1% of retailers say personalisation is strategically important, but only 10% have the mature capability to deliver it, and just 12% are confident their data is ready for AI. The ambition is racing ahead of the data foundation. That gap is where the money leaks.
Buying AI to run on data you do not trust is paying for a faster car with no idea where the road goes. The speed is not the problem. The blindness is.
There is a gap opening up in Australian retail and it is not a technology gap. It is a data gap. A new report from Arktic Fox, Six Degrees Executive and Amperity, drawn from more than 100 Australian marketing, digital, ecommerce and retail media leaders, found 89.1% of retailers say personalisation is strategically important. Only 10% say they have the mature capability to actually deliver it.
Worse, just 12% are fully confident their customer and product data foundations are ready for AI-led use cases. So the ambition is near universal and the readiness is rare.
Why it matters
AI does not resolve a mess. It automates the mess and scales it. Point an agent at fragmented customer data and weak definitions and you do not get personalisation, you get confident nonsense delivered at volume. The 10% who can deliver have not bought better AI. They have done the unglamorous work of getting their data in order first.
Share of Australian retailers with the mature capability to deliver the personalisation 89.1% of them call strategically important
This is the same pattern across the market, not just retail. The investment goes to the shiny layer because it demos well. The foundation gets skipped because it does not. Then the shiny layer underperforms and the AI gets blamed for a data problem.
What to do about it
Fix the foundation before you fund the agent. Clean identity, consistent definitions and reliable tracking are what make AI useful. Skip them and you are automating your worst data.
Audit what you actually hold. Most businesses cannot say with confidence which first-party data they have, where it sits and whether it is accurate. That audit is cheaper than any AI tool and worth more right now.
Start with one use case, not a transformation. Pick one customer signal you trust, act on it well, then widen. Learning to walk before you run applies to data as much as anything.
Measure data quality as a number, not a vibe. Track how complete and how fresh your customer records are. If you cannot put a figure on it, you cannot trust an agent to act on it.
The businesses that pull ahead in the next 18 months will not be the ones with the most AI. They will be the ones whose data is clean enough to let AI do something useful. The tool is available to everyone. The trustworthy data is the part most have not done.