![local voice logo [recovered]](https://mylocalvoice.io/wp-content/uploads/2026/10/Local-Voice-Logo-Recovered-250x56.png){kind=logo}
Local Voice research
77% of Customers Accept AI as Long as a Human Is Still Available
Invoca found 77% of buyers are more willing to use AI when they can easily reach a person. When AI fails, 68% hold the business responsible.
By Adam Brewer, founder of Local Voice. Figures checked October 10, 2026.
77% of buyers are more willing to use a brand's AI assistants when a person is also easy to reach, per Invoca's 2025 survey of 1,000 U.S. and U.K. consumers. When both are equally available, 67% still pick the person.
Invoca's 2026 follow-up shows what's at stake when the AI gets it wrong. When a brand's AI fails, 38% of buyers blame the brand alone and 14% blame the technology. Add the 30% who blame both, and 68% hold the business responsible, a Local Voice calculation.
Every figure was checked at its source on October 10, 2026. Derived figures are marked as Local Voice calculations, with the math shown in the methodology.
Key findings
77% of buyers are more willing to use a brand's AI assistants when they can also easily reach a human, per Invoca's 2025 survey of 1,000 U.S. and U.K. consumers.
When AI and human help were equally available, 67% of buyers preferred the human in Invoca's 2025 survey. In Invoca's 2026 report, 59% chose the human and 31% chose AI.
74% of buyers prefer AI when it resolves their issue faster than a human would, according to Invoca's 2025 survey of 1,000 U.S. and U.K. consumers.
When a brand's AI fails, 38% of buyers blame the brand alone and 14% blame the technology (Invoca, 2026). With the 30% who blame both, 68% hold the brand responsible (Local Voice calculation).
The share of buyers who said AI made their buying experience worse fell from 29% in 2025 to 18% in 2026, per Invoca. In 2026, 46% said AI made it better.
What does the 77% measure?
It measures willingness with a condition attached. Invoca found 77% of buyers are "more willing to use a brand's AI assistants" when a human is also easy to reach. Invoca's headline shortens that to 77% accepting AI when human support remains available.
The survey covered 1,000 consumers in the U.S. and U.K. who researched and made a high-stakes purchase in the prior 12 months. Sago Online Research fielded it, and Invoca published the results on July 24, 2025.
Home services was one of 7 purchase categories. The others were automotive, healthcare, financial services, insurance, telecom and travel. Invoca asked about AI assistants in general, so the 77% isn't specific to phone calls.
Figure 1. 77% of buyers are more willing to use AI when they can easily reach a person
Share of buyers agreeing with each statement. Invoca 2025, 1,000 U.S. and U.K. consumers who made a high-stakes purchase.
Four separate questions. The bars don't add up and shouldn't be combined.
Source: Invoca, B2C Buyer Experience Study, press release dated July 24, 2025.
Buyers lean toward a person when the options are equal: 67% prefer human help then. They lean toward AI when it's faster: 74% prefer AI when it resolves the issue sooner.
The 77% fits both answers. Most buyers will try the AI, as long as a person stays one step away.
Do customers still want to talk to a person?
Most do, when the choice is even. In Invoca's 2026 report, 59% of buyers chose the human when AI and human help were equally available. 31% chose AI, and Invoca didn't report the other 10%.
That 59% is 8 points below the 67% Invoca reported in 2025. The two surveys measured different people a year apart: 1,000 U.S. and U.K. consumers in 2025 and 693 buyers in 2026. Read the drop as a direction only.
Invoca's 2026 report advises connecting every AI touchpoint to a person the moment the buyer asks for one or is ready to buy. It doesn't define what counts as easy access to a human.
On a phone call, easy access takes two plain forms. During open hours, it's a live transfer to someone on your team. After hours, it's a callback time the caller can count on.
For the tasks callers are happy to leave with AI, see what customers will let AI do on the phone.
Who gets blamed when a business's AI fails?
The business does. Invoca's 2026 report found 38% of buyers blame the brand alone when its AI fails. Just 14% blame the AI technology, and 30% blame both equally.
Add the brand-only and shared groups, and 68% of buyers hold the business responsible, alone or in part. Invoca calls it "roughly two-thirds." Blame on the brand alone outweighs blame on the technology alone 2.7 to 1.
Figure 2. When a brand's AI fails, 68% of buyers hold the brand responsible
Who buyers blame when a brand's AI fails. Invoca 2026, 693 buyers.
- The brand alone38%
- The brand and the technology equally30%
- The AI technology alone14%
- Other or not reported18%
Invoca doesn't say what the remaining 18% answered.
Source: Invoca, 2026 B2C Buyer Experience Report, June 9, 2026. 68% and the other or not reported slice are Local Voice calculations.
Invoca's report says "The vendor takes none of the heat." If your phone agent books the wrong day or loops on a question, the caller remembers your company name. Treat the agent as part of your brand, the same as your trucks and your techs.
One bad experience costs less than it once did. Invoca reports the share of buyers who'd stop doing business with a company after one misstep fell from 66% in its prior report to 40%. That still leaves 4 in 10 buyers ready to walk.
Are buyers warming to AI?
Yes, by Invoca's measures. In its 2026 report, 18% of buyers said AI made their buying experience worse, down from 29% in 2025. 46% said AI made it better, and 37% said it made no difference.
Those three 2026 answers add to 101% because of rounding. Invoca says much of the drop in "worse" moved to "made no difference," which rose from 29% to 37%.
Table 1. What changed between Invoca's 2025 and 2026 buyer surveys (Local Voice calculation)
Invoca B2C Buyer Experience Study (July 2025; 1,000 U.S. and U.K. consumers) and 2026 B2C Buyer Experience Report (June 2026; 693 U.S. buyers). Human connection in 2026 is 98% in Invoca's blog and 83% in its press release, so no change is shown. The 2025 values for one misstep (66%) and no difference (29%) come from Invoca's 2026 report. Change is a Local Voice calculation. Sorted by size of change.
The 2025 and 2026 surveys reached different people at different times, so read the changes as direction. Fewer buyers said AI made things worse than a year earlier. Invoca's 2026 blog and press release disagree on human connection (98% and 83%), so we don't compare it with 2025.
For a phone agent, that means callers will give AI a fair hearing and still expect a person behind it. For how that fits with low comfort on voice calls, see why only 33% are comfortable with AI on a phone call.
What to do this week
Write the handoff first. Decide which calls go to a person and how fast they get there. Invoca's 2025 survey found 77% of buyers are more willing to use AI when a human is easy to reach. For a full setup checklist, see the AI phone setup checklist.
Make the agent sound like your shop. Load your service area and the answers your best dispatcher gives. In Invoca's 2026 data, 68% of buyers blame the brand, alone or in part, when its AI fails.
Call it yourself before customers do. Ask for a price, reschedule a job, report an emergency and ask for a person. If any call would embarrass you coming from a new hire, fix the script.
Read the transcripts every week. Flag every call where the caller asked for a person and didn't reach one. Invoca's 2026 report found 40% of buyers would stop doing business with a company after one misstep.
Ask any vendor to demo the transfer. Watch a call move from the AI to a person before you sign. That includes us: Local Voice sells AI phone agents.
Frequently asked questions
Do customers accept AI answering the phone?
Most will, with a person as backup. Invoca's 2025 survey of 1,000 U.S. and U.K. buyers found 77% are more willing to use a brand's AI when they can easily reach a human. That question covered AI in general. A separate Vonage survey of different people, in 2024, found only 33% comfortable with AI on voice calls.
Do people prefer a human or AI for customer service?
A human, when the two are equal. Invoca found 67% of buyers preferred human help in 2025 when both options were equally available, and 59% chose the human in 2026. Speed changes the answer: 74% prefer AI when it resolves their issue faster than a person would.
Who do customers blame when a company's AI gets it wrong?
Mostly the company. Invoca's 2026 report found 38% of buyers blame the brand alone when its AI fails, and 14% blame the technology. Another 30% blame both. Combined, 68% hold the brand responsible in whole or in part, a Local Voice calculation from Invoca's figures.
Are customers getting more comfortable with AI?
Yes, by Invoca's measures. The share of buyers who said AI made their buying experience worse fell from 29% in 2025 to 18% in 2026. 46% said it made the experience better. The two surveys reached different buyers a year apart, so read the change as a direction.
Is it true that 79% of callers can't tell an AI receptionist from a human?
We couldn't find a study behind it. In September 2026, AInora's statistics page credited the 79% figure to Stanford HAI. By October 10, 2026, that page listed it among figures it had removed. We found no Stanford HAI study by that name. Vonage's 2024 survey found 91% of consumers want AI use disclosed, so tell callers either way.
Free tool
What do missed calls cost your business?
Pick your trade, enter how many calls you miss and what a job is worth. The free calculator shows the sales you lose each month, each year and per missed call, using published booking and close rates.
Open the calculatorMethodology
Formulas
Share holding the brand responsible = brand alone + both equally = 38% + 30% = 68%. Blame ratio = 38% รท 14% = 2.7. Other or not reported = 100% minus 38% minus 30% minus 14% = 18%. Change in points = 2026 value minus 2025 value.
Invoca 2025 survey
B2C Buyer Experience Study, press release dated July 24, 2025. 1,000 consumers in the U.S. and U.K. who researched and made a high-stakes purchase in the prior 12 months, in automotive, healthcare, financial services, home services, insurance, telecommunications and travel. Fielded by Sago Online Research. Invoca doesn't publish field dates or question wording. The release body words the 77% as buyers being "more willing to use a brand's AI assistants" when they can also easily reach a human representative.
Invoca 2026 report
2026 B2C Buyer Experience Report, blog post by Owen Ray dated June 9, 2026. The page gives a sample of 693 buyers. Invoca's 2026 press release describes the figures as U.S. only, while the 2025 survey covered the U.S. and U.K. It doesn't publish fielding details, dates or question wording. It compares several figures with the prior year's report, including "made no difference" (29% vs 37%) and "stop doing business after one misstep" (66% to 40%). Invoca doesn't label which of 29% and 37% is 2026. We read 37% as 2026 because 46% + 37% + 18% = 101%, a rounding match.
Comparing the two years
The 2025 and 2026 surveys measured different people at different times, and Invoca publishes the question wording for neither. Table 1 pairs figures only where the published wording matches closely. Changes are in percentage points.
Unreported shares
Invoca reports 38%, 30% and 14% for who buyers blame. It doesn't describe the remaining 18%, so Figure 2 shows it as other or not reported. When AI and human help are equally available, 59% chose a human and 31% chose AI. The remaining 10% (100% minus 59% minus 31%) isn't reported.
Scope
Both Invoca surveys cover high-stakes purchases across industries, with home services one of several. They ask about AI assistants in general and don't isolate phone calls. Vonage's figures in the FAQ come from 1,000 U.S. adults who had recently used a brand's AI-based support, surveyed by Pollfish in September 2024.
Rounding
Percentages are shown as published. The blame ratio is rounded to one decimal, half up.
When we checked
Invoca's 2025 press release, Invoca's 2026 report page and Vonage's press release were checked at the source on October 10, 2026. In September 2026, the "Stanford HAI 79%" claim ran on AInora's statistics page. On October 10, 2026, that page listed it as removed. The September sighting comes from Local Voice's research notes.
Sources
Invoca. Invoca Study Reveals Consumers Embrace AI, but Reject Replacing Humans, in the Buying Experience. Press release, July 24, 2025. Accessed October 10, 2026.
Invoca. 2026 Report: Consumers Embrace AI, Demand for Fast Responses from Brands Increases. Owen Ray, June 9, 2026. Accessed October 10, 2026.
Vonage. Vonage Survey Reveals Increased Acceptance of AI in Customer Support, with Nearly Two Thirds of Consumers Admitting More Efficient Interactions. Press release via PR Newswire, November 21, 2024. Accessed October 10, 2026.
Stop missing calls
Answer every call, even after hours.
Local Voice answers your phone and books callers onto your calendar. Plans start at $397 a month for 500 minutes, plus a $500 setup fee. Call the demo line to hear it work.