Who's Lagging, and Why: A Forecast for AI Assistants in Australian Business
If banking and insurance are the early adopters of AI customer service, plenty of everyday Australia is the long tail. Our survey of ~500 brands found whole sectors where a real AI assistant is still rare or absent. The useful questions are why — and who moves next.
The laggards
At the bottom of the table, AI assistants barely register. In several sectors we found none at all.
Why they lag
The barriers aren't mainly technical — they're structural, and they differ by sector:
| Sector | Main barrier | Would AI help? |
|---|---|---|
| Dining / QSR | Franchise fragmentation — each store is its own operator | Limited (ordering apps already cover it) |
| Automotive | Dealer networks, not one central support desk | Yes — bookings, parts, recalls |
| Property | High-touch, agent-led, relationship sales | Yes — listing enquiries, inspections |
| Logistics | Thin margins, B2B-first mindset | Strongly — parcel tracking is a perfect fit |
| Energy retail | Has chat, but AI not switched on | Yes — billing, plan switching, outages |
| Health | Privacy, regulation, clinical risk | Cautiously — triage, bookings, not advice |
| Travel | Complex, exception-heavy enquiries | Yes — changes, refunds (Qantas already) |
Two patterns dominate. The first is fragmentation: dining, automotive and real estate are networks of independent franchisees, dealers and agencies, so there's no single website or support team to put an assistant on. The second is enquiry shape: property and travel lean on high-touch human relationships, while health is held back by privacy and clinical risk rather than cost or capability.
The near-term wave is already half-built
Here's the forecast's key insight. The brands most likely to adopt AI next are not the half with no chat at all — they'd have to roll out a widget first. They're the ~150 brands that already run a chat widget but haven't switched on AI. They have the integration, the vendor and the support team; they're one generative-AI tier away. And the major platforms (Zendesk, Salesforce, Intercom, Genesys) are shipping exactly those tiers now.
The forecast
Putting it together: adoption won't climb evenly — it will climb fastest where a widget already exists and a platform upgrade does the work, and slowest in the fragmented and high-touch sectors that lack the front door entirely.
These ranges extrapolate from the current floor, the chat-enabled headroom, and the pace of generative-AI rollouts by incumbent platforms. They're a considered forecast, not data — treat them as direction and magnitude, not precision.
Would AI actually help the laggards?
For some, clearly yes. Parcel tracking (logistics), billing and outages (energy), service bookings and recalls (auto) and refunds and changes (travel) are high-volume, rule-bound enquiries — the exact shape AI assistants handle best. For others — fine dining, boutique property, clinical health — the value is narrower, and a bot may be the wrong answer where the whole proposition is human attention.
The lesson from the leaders is that AI customer service pays off where enquiry volume is high and answers are repeatable. Plenty of laggards fit that description and simply haven't moved yet. That, more than any technology gap, is what the next two years will close.
See the full picture in the adoption survey, the stacks that will carry the upgrade in the tech-stack survey, and how we measured it all.