Published 2026-09-05 · Last updated 2026-09-05
Most customers will talk to an AI receptionist for routine tasks like booking a time or checking hours, and their comfort rises when the AI can actually finish the job. Salesforce found in April 2025 that 39 percent of US consumers are comfortable with AI agents scheduling appointments for them, and CallRail found in September 2025 that 50 percent would use AI to book an appointment, while 13 percent reject AI for every task. The dividing line is not the technology. It is whether the call is simple or hard.
How comfortable are customers, task by task?
Four separate surveys asked the question different ways, and the answers line up: the simpler the task, the more people will hand it to an AI.
| Task | Share willing | Source |
|---|---|---|
| Check a business's hours or location | 66 percent | CallRail, September 2025, 1,000 US consumers |
| Hand routine tasks to an AI assistant | 67 percent | Zendesk CX Trends 2025, about 5,100 consumers in 22 countries |
| Book an appointment | 50 percent | CallRail, September 2025 |
| Use an AI agent as a personal assistant | 44 percent | Salesforce, April 2025 |
| Let an AI agent schedule appointments | 39 percent comfortable | Salesforce, April 2025 |
| Work with an AI instead of a person to avoid repeating themselves | 34 percent | Salesforce, April 2025 |
| Reject AI for every task | 13 percent | CallRail, September 2025 |
Zendesk's report, published in November 2024, also found half of consumers had already engaged with voice AI. For a trades business, the calls that matter most are the simple ones: 'Do you do water heaters?' 'Are you open Saturday?' 'Can someone come Tuesday?' Those are exactly the tasks where willingness is highest.
Does age change the answer?
Yes, and mostly in one direction. CallRail found 91 percent of Gen Z would use AI to book an appointment against 50 percent overall, and Salesforce found 70 percent of Gen Z would use an AI agent as a personal assistant against 44 percent overall. Younger homeowners are not just tolerating the AI. Many of them expect it.
Familiarity is spreading across every age group. Pew Research Center surveyed 5,119 US adults in February 2026 and found about half use AI chatbots, one in four use one daily, and 44 percent use ChatGPT. A caller who used an AI assistant this morning is not startled to hear one answer your phone.
The condition that decides it: can the AI finish the call?
A NewtonX study for Ada, run from December 2025 to February 2026 with 2,000 consumers who had used AI customer service in the prior year, found 59 percent prefer instant, 24/7 AI service over waiting for a human, but only when the AI can actually resolve the issue. Ada commissioned the study, so weigh it as vendor research. The pattern still holds in the other data.
Zendesk found 64 percent of consumers trust AI agents more when they show friendliness and empathy, and CallRail's respondents reported 42 percent favorable AI interactions against 26 percent negative. The favorable experiences are the ones where the caller got what they came for. The negative ones are the loops, the misheard names, and the 'let me transfer you' that goes nowhere.
What are businesses seeing on their side?
Adoption in customer service jumped in a single year. Salesforce's State of Service report, surveying 3,075 service professionals in March and April 2026, found the share of service organizations using AI agents rose from 39 percent in 2025 to 66 percent in 2026, 70 percent of adopters saw measurable value within 60 days, and customer satisfaction was the top improved measure. In the same report, 85 percent of service organizations use at least one form of AI.
Gartner predicted in March 2025 that agentic AI will autonomously resolve 80 percent of common customer service issues by 2029 and cut operational costs by 30 percent. Customers notice the coverage too. CallRail found 52 percent see a business offering an after-hours AI assistant as a sign of superior service.
Where does a human still need to pick up?
The surveys draw the line for you. Willingness is high on simple, repeatable tasks and drops on anything with stakes or emotion. A good AI receptionist is set up to know the difference and get out of the way.
- Emergencies: read the approved safety instruction and transfer to the person on call
- Upset callers and complaints: acknowledge, record the issue, and alert the owner
- Pricing outside the approved list: take details and promise a callback with a stated time
- Anything the business has not approved it to say: admit it cannot confirm and hand off
Honesty is part of the design, not a disclaimer. The AI says what it is in the first sentence. Callers judge it on whether it helps, and the data says most of them will give it the chance.
Common questions
Will older customers hang up on an AI receptionist?
Some will, and a transfer path should exist for them. But willingness is not confined to the young: CallRail found only 13 percent of consumers reject AI for every task.
What makes the difference between a good and a bad AI call?
Resolution. The Ada study found consumers prefer AI only when it can actually resolve the issue. An AI that can book, confirm, and transfer earns the preference. One that only takes messages does not.
Should the AI pretend to be a person?
No. Disclose it in the first sentence. Familiarity with AI assistants is high, and a caller who discovers the truth later trusts the business less, not the AI.
