Cold Calling Beats AI Email 6 to 1. What B2B Teams Should Fix First

New research released September 1, 2026 by ValueSelling Associates and Selling Power finds that sellers rate human cold calls six times more effective than AI-assisted personalized email for reaching new prospects: 47% versus 8%. Only client referrals, at 74%, ranked higher. If you fund an outbound stack, this stat is about to show up in every sales training pitch you see this quarter, so it is worth knowing what the study actually measured, where it is shaky, and what to change in your own prospecting engine. Here is the breakdown.

Read next: What is an AI SDR, and what should it actually do?

What did the ValueSelling research actually find?

ValueSelling Associates and Selling Power surveyed 153 sales leaders and reps across industries on which tactics earn time on a decision maker's calendar. The effectiveness ranking, top to bottom: client referrals (74%), cold calls by sales reps (47%), LinkedIn and social selling (40%), conferences and trade shows (30%), partner or ecosystem introductions (26%), email outreach campaigns (24%), and AI-assisted personalized email outreach at 8%, dead last.

That last gap is the headline. Every human-led tactic outranked every automated one, and the most automated tactic of all, the AI-personalized email that half the sales tech industry is currently selling, was rated effective by fewer than 1 in 12 sellers.

The study also reran questions from a 2018 edition, which makes the comparison unusually useful. Eight years of aggressive sales tech adoption changed the ranking almost not at all.

Meanwhile execution is poor: only 19% of sales leaders rate their reps' cold calling as excellent or very good, 31% say good, and 50% say fair or poor. ValueSelling CEO Julie Thomas put it plainly: "cold calling isn't dead." Her read is that the issue is not the method but the skill, and the fix is "to instill seller confidence through sales training."

Should you trust the 6-to-1 number?

Treat it carefully. This is 153 sellers rating perceived effectiveness, not a conversion study of measured meetings booked per channel. And ValueSelling sells sales training, so a finding that the fix is training is a finding that pays the researcher. Neither of those makes the data wrong, but they mean you should read it as a strong directional signal, not a law of physics.

The direction, though, matches what we see in client outbound stacks every week. Pure-AI email at volume gets ignored, and the teams booking meetings are the ones using AI where it is quiet (research, list building, call prep) while a human does the part the buyer actually experiences. Notably, sellers in this same study are leaning on AI hard: 55% use it to draft prospecting emails, 55% to research prospects, 46% to personalize outreach, and 43% to build call-prep notes. The tooling is not the problem. Pointing all of it at drafting more email is.

Why is AI-written email rated so low?

Because everyone has it. When personalization is generated, it stops signaling effort, and effort was the thing personalization used to signal. A buyer who receives forty "noticed you just raised a Series B" emails a week learns to delete all forty. The 24% rating for plain email campaigns versus 8% for AI-personalized outreach is the study's quietest and most damning detail: sellers rate the un-personalized version three times higher, presumably because at least it is honest about being a campaign.

The phone works for the mirror-image reason. Almost nobody calls anymore, so a competent human call is scarce, and scarcity gets attention. The same study shows fear of calling is falling (48% of leaders cited it in 2018, 39% now), which suggests the reps who remain on the phone are getting more comfortable while the crowd thins around them.

What should demand gen teams do now?

Keep AI, move it upstream. Use it for account research, trigger detection, and call preparation, the jobs sellers in this study already use it for. That is exactly how we wire an AI SDR prospecting engine: the system builds the list, the research, and the brief, and your reps man the seat.

Put call blocks back on the calendar. Half of leaders rate their teams' calling as fair or poor, which reads as opportunity: the bar for a good call is on the floor. If the 47% effectiveness rating holds even half true against your funnel, a daily call block on your warmest accounts outperforms another sequencing tool.

Fund the referral and partner motion first. The top of the ranking is client referrals at 74% and partner introductions at 26%, and both are chronically under-resourced next to email tooling. A systematic referral ask after every closed-won and every QBR costs nothing and beats every cold channel in this dataset.

Frequently asked questions

Is cold calling dead in 2026?

No. In ValueSelling Associates' September 2026 research, cold calls rank as the second most effective way to reach new prospects at 47%, behind only client referrals at 74%, and the rating is one point higher than it was in 2018. What has changed is execution: half of sales leaders rate their reps' calling as fair or poor.

How much more effective is cold calling than AI-written email?

Six times, by seller rating: 47% of the 153 sellers surveyed rated cold calls effective for reaching new prospects versus 8% for AI-assisted personalized email. Note this measures perceived effectiveness in a survey, not booked-meeting conversion data.

What is the most effective way to reach new B2B prospects in 2026?

Client referrals, rated effective by 74% of sellers, followed by cold calls (47%), LinkedIn and social selling (40%), conferences (30%), partner introductions (26%), email campaigns (24%), and AI-personalized email (8%).

Should B2B teams stop using AI for outbound?

No, they should move it upstream. Sellers in the study already use AI for prospect research (55%), drafting (55%), personalization (46%), and call prep (43%). The data argues against AI writing the touch the buyer sees, not against AI running the research and preparation behind a human call.

Sources: ValueSelling Associates press release, September 1, 2026 and Demand Gen Report coverage. All figures verified against the primary release on September 1, 2026.