The cold call isn't dead. The bad LinkedIn DM is.

By Marcus Webb, Outbound Strategy Last updated: 2026-05-28

---

A short list of things you have probably read in the last 12 months:

• "Cold calling is dead."
• "LinkedIn is the new outbound."
• "Email is the new outbound."
• "Actually, the phone is back."

Each of those headlines was written by a vendor whose product happens to be the answer. Strip the marketing and the data tells a calmer story: every legitimate outbound channel still works for someone. What has changed is the cost of doing any of them badly. The bad LinkedIn DM, with its Hi {firstName} and its generic ask, is genuinely dying. The cold call is not.

This post argues a contrarian read on the "cold call is dead" meme, anchored to first-hand Reachium platform data on LinkedIn reply-rate decline and the public cold-call benchmarks that almost nobody linking the obituary bothers to cite.

---

Where did "the cold call is dead" headline come from?

Mostly from the 2019 to 2023 wave of LinkedIn automation vendors. The pitch was clean: dials are slow, gatekeepers are brutal, mobile carriers label your number as spam, so move budget to LinkedIn where you can touch 100 prospects an hour from a browser extension. There was a real insight inside it. Mobile spam labeling did make dial economics harder, and named-account selling did need a second channel.

The framing went too far. "Harder" became "dead." A channel that still books meetings every day got buried under a category of think-pieces written by people selling the alternative. The same playbook is now being run in reverse: vendors selling AI dialers are quietly publishing pieces about how LinkedIn is saturated. Both camps are half right and fully self-interested.

The honest read is that the phone never became un-pickable. The cold DM, by contrast, has a hard cap on how badly you can do it, because LinkedIn can throttle, restrict, or eventually revoke the account that sends too many low-quality requests. Stack those two facts and the obituary starts to look upside down.

Is the cold call actually dead?

No, and the public data is not subtle about it. The Bridge Group's long-running SDR benchmark, the most-cited dataset in B2B sales development, still places multi-channel SDRs at roughly 50 dials and 65 emails and 20 to 25 LinkedIn touches per day, with about 4.4 connects per 100 touches when measured across all activity. SalesHive's 2025 cold-calling benchmark for B2B teams puts the live-connect rate higher, in the 16% to 22% band for trained reps, while Gong's analysis of more than 300 million calls landed the average connect at around 5.4% and the top quartile at 13.3%. The numbers vary by definition (dial-to-connect vs touch-to-connect), but the floor never falls to zero.

The conversion side is the part the obituary skips. When a B2B decision-maker does pick up the phone, the conversation is the highest-bandwidth touch in outbound. Thirty seconds of voice beats two weeks of "circling back" by inbox. The averages reported by SalesHive in 2025 put cold-call meeting conversion at roughly 2.3% to 2.5% per dial for typical teams and 5% to 8% for trained ones, which is comparable to or above the meeting conversion of cold email for the same effort.

The cold call is alive because two things are still true. Nobody can ban a phone number for trying. And a connected call still books the meeting at a rate the inbox cannot match.

Why are LinkedIn DM reply rates falling then?

Because the channel has been flooded with templates that all look the same. Reachium's data across 316,703 outreach sequences run on the verified LinkedIn API shows the reply rate of accepted requests drifting from roughly 26% to 34% in the second half of 2025 down to roughly 16% to 26% in 2026, a decline of about ten percentage points at the midpoint. Acceptance rates held steadier, around 25% to 30%, which means the falling number is not "can I get connected" but "did the message I sent after connecting deserve a reply."

That is a quality story, not a channel story. The recipients did not stop using LinkedIn. They stopped replying to the version of LinkedIn where every message reads "Hi {firstName}, saw you're in {industry}, would love to {generic ask}."

Add the volume tax on top. Reachium's data shows acceptance peaking at 34% for accounts sending 10 to 19 invites a day and falling to 30.6% for accounts pushing 20 to 29 a day. The platform caps the upper end on purpose, but even within the safe band, more volume produces fewer accepts. The harder you mash the button, the less it works. That is the bad DM losing in real time, not the channel.

Is cold email or LinkedIn DM the one that is actually dying?

Both are getting harder for the same reason, which is buyer fatigue with templated outreach, and neither is dying in the sense the headlines mean. A useful frame: cold email has a deliverability cliff (warm-up, domain reputation, spam filters) and LinkedIn has an account-safety cliff (rate limits, restrictions, the public bans that hit browser-automation tools like the HeyReach incident in March 2026). Both penalize the spammer faster than they used to.

What that punishes is volume without context. What it rewards is a sequence where each touch carries a real reason it exists: a hiring signal, a funding event, a piece of content the prospect engaged with, a peer they share. Channels are not the variable. Effort is. The post on the common LinkedIn outreach mistakes that kill reply rates walks the diagnostic for the LinkedIn half. The cold-call half lives in the public Bridge Group and Gong datasets, and they tell the same story: top quartile is two to three times the average because top reps prepare and average reps read a script.

Should sales leaders still hire SDRs to dial in 2026?

For the right motion, yes. For the wrong motion, no. The honest steel-man for the "cold call is dead" view holds up in two cases: low average contract value where a 100-dial close cannot pay for itself, and audiences who genuinely will not pick up an unknown number (a portion of younger ICPs, certain functional roles). In both, dial economics break before the conversation starts.

For high-ACV named-account selling, the phone is still the bandwidth king. The right test is not "is the cold call dead" but "what is one connected call with a director-level buyer worth to my pipeline, and how many dials does it take to get one." If the answer is "more than the cost of the SDR-hour," the phone stays in the stack. If it isn't, the budget should move to a personalized LinkedIn motion plus signal-based email, not to a louder version of the same script.

What is the right multi-channel mix in 2026?

The mix that is winning right now in the data is small and disciplined: targeted phone for the top tier of the account list, where each connect is worth the dial, plus personalized LinkedIn DM sequences for the broader ICP, both fed by signals (job changes, hiring, funding, content engagement, mutual peers). Cold email runs underneath as the catch-all with deliverability hygiene. Nothing in that stack is "100 untouched dials" or "200 mail-merge invites."

The LinkedIn half is the part that gets bungled most. Acceptance and reply only hold up if the first message references a real reason the outreach exists. Reachium runs personalized sequences on LinkedIn's verified API with AI Personalization writing the first line per prospect rather than a merge field, and it calibrates invite volume to the 10 to 25 per day band where acceptance is highest. A framework for what to actually post on LinkedIn and a guide on how to personalize LinkedIn outreach at scale cover the messaging side; the volume-tax write-up and the 2026 outreach benchmarks cover the math. The pieces are there. The discipline is the bottleneck.

FAQ

Should we still hire SDRs to dial in 2026?

For high-ACV, named-account motions, yes, because connected calls book meetings at rates the inbox cannot match. For low-ACV, high-volume motions, the math has broken, and budget belongs in personalized LinkedIn and signal-based email instead.

Is cold email also dying?

It is getting harder for the same reason LinkedIn DM is, which is buyer fatigue with templates. Deliverability infrastructure (warm-up, domain reputation) now punishes volume-first senders faster than it used to. The fix is the same: fewer, more contextual touches.

What counts as a "personalized enough" LinkedIn DM?

A first message that references something true about the recipient that a merge field cannot produce: a recent role change, a piece of content they posted, a mutual customer, a hiring signal. Reachium's AI Personalization writes that first line per prospect rather than templating it, which is the operational form of "personalized enough."

Do video or voice DMs bring LinkedIn back?

They lift reply rates when used sparingly to senior buyers, because they prove a human was on the other side. They die fast when scaled to every prospect. Treat them as the top-of-list move, not the default.

How do we measure phone, LinkedIn, and email as one motion?

By account, not by channel. A booked meeting that took two dials, one DM, and one email should credit the account-level sequence, not fight over which channel "won." Track meetings booked per account contacted, not requests sent or dials made.

Sources

• Linked Insider: LinkedIn outreach mistakes that kill reply rate
• Linked Insider: the LinkedIn volume tax
• Linked Insider: 2026 LinkedIn outreach benchmarks
• Reachium
• The Bridge Group: SDR Metrics and Compensation Report
• SalesHive: Cold Calling Benchmarks for B2B Sales Teams in 2025
• Gradient Works: 2024 B2B sales benchmarks