Twilio Study Finds Australians' Digital Patience Waning As AI Reshapes Customer Service
Australians are famous for a relaxed attitude. In customer service, that patience is starting to crack-especially online.
Twilio's Decoding Digital Patience study, run with YouGov in August 2025, surveyed 7,331 consumers across APAC, including 1,031 in Australia. The findings, released alongside Twilio's SIGNAL event in Sydney, show a widening gap between what customers expect and what AI-powered systems deliver.
The patience gap is real
- 96% of Australians believe they're expected to be patient and polite during support interactions.
- Only 64% say they actually remain patient when dealing with brands online.
- 50% say AI in customer service makes them less patient, and they're more likely to lose patience with AI than with a human.
Where AI falls short
- Perceived patience with AI chatbots: 53%. With automated voice menus (IVR): 54%.
- Patience jumps with humans: 88% for live chat, 90% for phone calls.
- Top frustrations:
- AI not getting the question: 52%.
- Repeating themselves multiple times: 48%.
- Scripted or robotic answers: 46%.
Human touch still matters
- 78% have used AI tools like chatbots or voice assistants; 39% are satisfied.
- 49% prefer to start with a human agent (vs 42% across APJ), even if it takes longer.
- 84% say easy transfer from AI to a human agent is important.
"Australian consumers certainly see themselves as more patient than the average respondent across the rest of the APJ region. But that patience evaporates quickly when service lacks the human touch. The real differentiator will be how well brands strike that fine balance between efficiency with genuine, human connections that build enduring trust, loyalty and long-term competitive advantage," said Nicholas Kontopoulos, Vice President of Marketing for Asia Pacific and Japan at Twilio.
What it costs to get automation wrong
- 20% will stop using the product or switch to a competitor if automated interactions take too long.
- 34% are more likely than their APJ peers to downgrade their opinion of a brand after poor service (vs 29% regionally).
- 32% will share negative experiences (vs 25% regionally).
- 40% regret starting long, complicated online sign-ups for new services.
What customers value most
- Quick resolution: 49%.
- Clear, easy-to-follow instructions: 43%.
- Confidence that their personal data is safe: 38%.
For privacy expectations in Australia, see the Australian Privacy Principles from the OAIC. Learn more.
Playbook for customer support leaders
Design for human-first, AI-assisted support
- Offer opt-out and fast escalation to a human within two steps. Make the transfer seamless, with full context passed through.
- Route by intent and value: high-friction or high-stakes issues go to skilled agents immediately.
Fix the basics customers notice
- Compress time-to-resolution. Track first contact resolution, queue time, and containment without repeat contacts.
- Cut repetition with conversation memory. Auto-summarize and share context across channels and handoffs.
- Replace canned scripts with dynamic responses grounded in customer data and past interactions.
Make AI actually useful
- Train models on real transcripts and local phrasing. Continuously test against top failure reasons: intent miss, repetition, robotic tone.
- Give AI narrow, high-confidence tasks first: authentication, status checks, simple updates, proactive reminders.
- Set confidence thresholds. When uncertain, escalate-not guess.
Simplify journeys that cause regret
- Shorten forms with progressive steps and autofill. Save progress and allow resume across devices.
- Explain why you need each data point. Show security badges and plain-language privacy notes near inputs, not buried in footers.
Measure the patience curve
- Instrument drop-off points in IVR, chatbots, and forms. Correlate with sentiment and repeat contact rate.
- Review escalations where customers typed "agent," "human," or used negative language. Prioritize those intents in quarterly tuning.
Coach agents for the moments that matter
- Equip agents with real-time suggestions and next-best actions drawn from customer context.
- Reward outcomes customers care about: clarity, speed, and empathy. Scorecards should reflect that mix.
Bottom line
Australians will wait for people, not for poorly designed automation. Let AI remove friction, not feelings. Make escalation obvious, keep instructions clear, protect data, and measure where patience runs out-then fix those moments fast.
Want structured training for AI-assisted support workflows and agent skills? Explore curated programs for support roles at Complete AI Training or see the latest AI courses.
Your membership also unlocks:
