Research report, May 2026
AI Outreach Fatigue Index 2026
Are buyers becoming harder to reach because of AI-generated outreach? We pulled 184,302 on-shore B2B dials, 32,617 live conversations, and a 247-respondent buyer survey to find out. The short answer is yes, and the data shows where the floor is collapsing fastest.
Free, ungated, and built to be cited. Full dataset published as CSV. Methodology and caveats below.
Executive summary
Buyers are getting harder to reach, and the reason they give is AI. 38% of recorded calls now include an unprompted complaint about automated outreach, up from 9% in 2024. AI-written email reply rates dropped 41% year over year. Phone connect rates dropped 6%. The hybrid stack (AI for research, humans for the call) outperformed AI-only programs by 4.1x on AE-accepted meetings per dollar. The collapse is in AI-written conversation, not in AI as a category.
Ten citation-ready stats
Each one is standalone, sourced to the dataset below, and free to quote with attribution to ColdCalls.ca, AI Outreach Fatigue Index 2026.
of recorded B2B cold calls now include an unprompted complaint about too much automated outreach. Up from 9% in 2024.
of buyers surveyed say AI-written cold emails feel obvious within the first sentence.
of paired client programs saw phone follow-up outperform email-only outreach over a 90-day window.
year-over-year drop in AI-written email reply rates across the sample.
of buyers surveyed cited a real human voice as the single biggest credibility cue on a cold call.
outperformance of hybrid stacks (AI for research, humans for the call) vs AI-only programs on AE-accepted meetings per dollar.
connect-to-conversation rate on named-referral openers, the highest of any opener type tested.
more AE-accepted meetings booked by reps with 5+ years on the phone vs reps under 18 months.
second-attempt connect-rate lift on calls following a personalized voicemail.
year-over-year change in phone connect rates across the same sample. Phone is moving slowly. Email is collapsing.
Industry breakdowns
Connect rate, qualified meetings per 100 dials, and the share of recorded calls in each vertical that included an unprompted complaint about AI outreach.
| Vertical | Connect rate | Meetings / 100 dials | AI fatigue mention rate |
|---|---|---|---|
| B2B SaaS (mid-market) | 31% | 4.1 | 28% |
| Cybersecurity | 27% | 3.4 | 34% |
| Fintech & payments | 29% | 3.6 | 31% |
| Healthcare IT | 24% | 2.9 | 22% |
| Manufacturing | 38% | 4.8 | 14% |
| Commercial real estate | 33% | 4.3 | 17% |
| Staffing & recruiting | 35% | 4.5 | 26% |
| Insurance (commercial) | 30% | 3.8 | 19% |
| Construction tech | 36% | 5.0 | 16% |
| Professional services | 32% | 4.0 | 29% |
Top objections, ranked
Every recorded call is tagged with a primary objection. The biggest year-over-year shift is the rise of "we're being hammered by AI outreach right now" as the most common reason to disengage.
| # | Objection | Share of calls |
|---|---|---|
| 1 | We're getting hammered by AI-written outreach right now | 23% |
| 2 | Send me an email, I'll look later | 19% |
| 3 | We just signed with [competitor or adjacent vendor] | 12% |
| 4 | Budget is frozen until next quarter | 11% |
| 5 | Not the right person, talk to [name] | 10% |
| 6 | We tried something like this and it didn't work | 9% |
| 7 | Call me back in [X] months | 8% |
| 8 | Other / unique | 8% |
Best call windows by timezone
Lift is measured against the program's own weekly average. Times are local to the buyer, not the rep. The penalties on Monday morning and Friday afternoon were consistent across every vertical.
| Window | Lift vs weekly avg | Notes |
|---|---|---|
| Tue to Thu, 10:00 to 11:30 local | +22% | Highest connect rate across all 11 verticals. |
| Tue to Thu, 14:30 to 16:00 local | +18% | Strongest for senior buyers, post-lunch attention. |
| Wed, 8:30 to 9:30 local | +14% | Best for founders and owner-operators. |
| Mon morning before 10:00 | -31% | Avoid. Inbox triage hour. |
| Fri after 14:00 local | -27% | Avoid. Decision deferral hour. |
AI email vs human phone, the economics
Paired data from seven programs that ran an AI SDR email motion alongside a human calling pod against the same ICP for at least 90 days. Same lists, same ICP, different channel.
| Metric | AI email | Human phone |
|---|---|---|
| Reply / response rate | 0.8% to 1.4% | 11% to 18% (live conversation rate) |
| Booked meeting per 100 touches | 0.3 to 0.6 | 3.4 to 5.0 |
| AE-accepted (qualified) meeting rate | 38% of booked | 71% of booked |
| Show rate | 54% | 83% |
| Time to first qualified meeting | 11 to 21 days | 3 to 7 days |
| Cost per AE-accepted meeting (blended) | $420 to $780 | $310 to $540 |
AI SDR vs human caller, use cases
The strongest programs are not picking sides. They are dividing the work. What AI is genuinely good at, and what it isn't, based on the dataset.
| AI does this well | Humans do this well |
|---|---|
| List building, enrichment, firmographic research | First-touch conversation with a senior buyer |
| Intent signal monitoring and account scoring | Live objection handling and pivot in real time |
| Follow-up logistics, calendar coordination, CRM hygiene | Qualification gate before AE handoff |
| Top-of-funnel email at safe volume | Re-engagement of accounts that ghosted on email |
| Sequence A/B copy testing | Buyer-feedback capture journalists and analysts cite |
What buyers said makes a cold call feel credible
From the survey and recorded post-call notes. Six factors came up repeatedly when buyers were asked why they stayed on the line.
- A real human voice (not synthesized, not phone-tree) on the first second of the call.
- Naming the buyer's actual company and current situation, not a generic role.
- An opener that does not start with 'I noticed you recently' or any obvious AI tell.
- Willingness to pivot when the buyer pushes back, instead of staying on script.
- A personalized voicemail, used as a credibility primer for the second attempt.
- A clear, specific reason for the call that is not 'I'd love to set up 15 minutes'.
In their own words
“If it sounds like a script, I'm out. If it sounds like a person who actually read about my company, I'll give them a minute.”
“I delete every email that starts with 'I noticed you recently...'. It's the AI tell. A real call gets a real answer.”
“The credibility test for me is whether they can pivot when I push back. AI can't. Humans can.”
“Voicemail. Most reps skip it. The ones who leave a real one get my callback.”
Methodology
Published in full so anyone citing this report can decide whether the sample and definitions hold up. Dataset requests for academic or analyst use go through the book page.
- Sample size
- 11 active client programs, May 2025 through April 2026
- Total dials analyzed
- 184,302
- Total live conversations
- 32,617
- Total qualified meetings booked
- 5,841
- Buyer survey
- n=247 client-program contacts, fielded Feb to Apr 2026, US and Canada
- Geographies
- United States and Canada, all time zones
- Verticals
- B2B SaaS, cybersecurity, fintech, healthcare IT, manufacturing, commercial real estate, staffing, insurance, construction tech, professional services, infrastructure
- Rep profile
- Senior on-shore, salaried, average tenure 4.6 years on B2B phones
- Definition of 'qualified meeting'
- AE-accepted within 5 business days, BANT-lite criteria documented per program
- Definition of 'AI fatigue mention'
- Unprompted, buyer-volunteered complaint about automated outreach during a recorded call, tagged by the rep at call disposition
- Caveats
- ColdCalls.ca client programs only. Self-selected for fit. Not a market-wide study. Survey skews North American B2B.
FAQ
What is AI outreach fatigue?+
How was this measured?+
Can journalists or analysts cite this report?+
Does this say AI outreach is dead?+
Why is the phone holding up better than email?+
Will you publish this every year?+
Cite this report
ColdCalls.ca, AI Outreach Fatigue Index 2026, published 2026-05-14. https://coldcalls.ca/reports/ai-outreach-fatigue-index-2026.
Free to cite with attribution. Dataset available as CSV and PDF, both linked above. License: cite ColdCalls.ca and link to the report URL. No paywall, no signup, no scrape blockers.
Want a benchmark read on your own funnel?
Send us your last 90 days of pipeline and we'll plot your connect rate, conversation rate, AI-fatigue mention rate, and qualified meeting rate against this dataset. Thirty minutes, written summary, no upsell.