GuidesResearch · 2026
Australian business call statistics for 2026
We took the calls our AI receptionists answered for Australian businesses over a 90-day window in 2026, and measured when they arrived and how they ended.
Call 1800 570 693 to talk to the demo agent. Plans start at A$299 a month, GST included.
Findings
- 28%
28 per cent of calls to Australian businesses with an AI receptionist arrived outside weekday business hours.
Business hours run 9am to 5pm, Monday to Friday, in the business's local time. Weekends count as outside hours. We average each business's own share, so each business counts once.
- 13%
Callers rang Australian businesses more after 5pm on weekdays (13 per cent of calls) than across the whole weekend (8 per cent).
After 5pm means 5pm to midnight, Monday to Friday. The weekend means all of Saturday and Sunday. Both use the business's local time.
- 24%
24 per cent of calls answered by AI receptionists for Australian businesses ended before the caller said a word.
The caller said nothing the receptionist could transcribe before the call ended. We count hang-ups during the greeting and silent lines here.
- 23%
AI receptionists took a message for the business on 23 per cent of calls.
The receptionist recorded a message for the business. A call where the receptionist answered the question itself does not count.
- 14%
AI receptionists transferred 14 per cent of calls to a person at the business.
The receptionist put the call through to a number the business chose, such as the owner's mobile.
All published figures
| Figure | Share | Definition |
|---|---|---|
| Outside weekday business hours | 28.0% | Outside 9am to 5pm, Monday to Friday, local time |
| Weekday, before 9am | 7.6% | Midnight to 8:59am, Monday to Friday |
| Weekday, 9am to 5pm | 72.0% | 9am to 4:59pm, Monday to Friday |
| Weekday, after 5pm | 12.9% | 5pm to midnight, Monday to Friday |
| Saturday | 4.0% | All day Saturday |
| Sunday | 3.5% | All day Sunday |
| Caller said nothing | 23.9% | The call ended before the caller said anything the receptionist could transcribe |
| Caller hung up within 10 seconds | 8.2% | Out of calls with a metered length |
| Repeat call | 37.4% | The same number had already rung the same business earlier in the window |
| Message taken | 22.5% | The receptionist recorded a message for the business |
| Transferred to a person | 13.9% | The receptionist put the call through to a number the business chose |
The weekday, Saturday and Sunday rows split every call once.
Method
We ran one read-only query against the production database behind our receptionists. It returns averages and shares. No transcript, recording, phone number or business name left the database.
| Item | Detail |
|---|---|
| Sample | Inbound phone calls answered by the AI receptionists we run for Australian businesses |
| Window | A 90-day window in 2026 |
| Weighting | We count each business once and average each business's own share, so a busy line carries the same weight as a quiet one. |
| Local time | The timezone set on each business's account. Public holidays count as ordinary weekdays. |
| Publication rule | At least 10 businesses and 1,000 calls sit behind every figure, and removing any one business moves a figure by 5 percentage points or less. |
| Rounding | Whole numbers in the findings, one decimal place in the table |
| Not recorded | How long a call rang before the receptionist answered |
Excluded calls
We removed these calls before counting anything.
- Test calls, evaluation runs and simulated callers
- Calls through a website widget, and text chats
- Calls from staff and test handsets, and from any number that rang more than one business on the platform
- Calls to internal, demo and fixture lines, including the public demo line
- Calls the platform placed itself, such as the set-up call to a new customer
- Calls to accounts with no member outside our staff, which hold demos for prospects
Limits
Each business chooses which calls reach its receptionist, from every call to after-hours calls alone. The figures describe the calls the receptionists answered, which can be fewer than the calls a business received.
The businesses in the sample chose to use an AI receptionist. They are not a random sample of Australian businesses.
A business that kept the default timezone on its account has its calls placed in Sydney time.
Suppressed figures
These figures failed the publication rule, so the page gives no value for them.
| Figure | Reason |
|---|---|
| Median call length | Removing one business moved it too far |
| Calls that made a booking | Removing one business moved it too far |
| Spam, new enquiries and urgent calls | Too few businesses had calls classified by the post-call analysis |
| Callers who rang more than once | Too few callers sit behind it |
| Hour-by-hour and single-weekday splits | One band failed the rule, and the other bands would reveal it by subtraction |
| Split by industry | Too few businesses in each industry |
Citation and licence
Vertical AI (2026). Australian business call statistics for 2026. https://verticalai.com.au/research/business-call-outcomes
We license these figures under CC BY 4.0. You can reuse them in articles, reports and lists when you credit Vertical AI and link to this page.
The receptionists in this data are the same product as our demo line. Call 1800 570 693 to hear one, or book a call to talk about your own line.
Related pages
Get started
Tell us what your callers ring about, and we build and test the agent for you.