Cold email vs LinkedIn DMs, same client, same message: 335 emails delivered and zero calls versus 139 DMs and 22 calls booked. Every number inside.

"Should we do cold email or LinkedIn? Our SDR wants to fire up Apollo. Our investor says LinkedIn. We don't have budget for both."
I get this question on almost every strategy call.
Most answers you'll find are opinions. This one isn't. I ran cold email and LinkedIn DMs side by side for the same client, targeting the same ICP, with the same messaging, at the same time. One channel booked 22 sales calls. The other booked zero.
Here are the numbers, what happened, and the LinkedIn outreach strategy that produced the 22.
The client is an AI startup with a narrow target market. I recommended LinkedIn DMs. They convinced me to layer cold email on top. Fine — let's find out. So I:
The message was the same. The offer was the same. The only variable was the channel.
Here's what cold email produced:
Look at that open rate. 63.6% is a good open rate. People opened the emails. Then they did nothing, or they hit unsubscribe.
The replies we got:
When I say we got nothing from cold email, I mean we really got nothing. And a 12.5% opt-out rate is absolutely insane. One in eight people who received the email went out of their way to tell us never to email them again.

Same client, same window, same messaging:
That's a 20% reply rate and a 78.6% response-to-call rate. Of the people who replied, nearly 4 out of 5 booked a meeting.
The acceptance rate and reply rate were lower than my usual benchmarks because this client's market is so narrow that we had to send requests to people who aren't very active on LinkedIn. But that same narrow market is why the response-to-call rate was through the roof. When the right person replies, they're ready to talk. The full case study, with screenshots, is here.
Be honest: if someone gave you the choice between cold emailing 390 prospects for zero results or DMing 139 prospects and booking 22 calls, which would you choose?
The first email in our cold sequence didn't sell anything. It was a soft, human opener. And we still got "I am not interested but thanks for reaching out."
These are involuntary responses. The moment someone even suspects it might be a cold email selling something, they pull up the drawbridge and tell you to go pound sand. People have been swimming in cold email for years, and AI made it worse. No one wants your cold email. They dread it.
Now compare that to the same exact messaging landing in a LinkedIn inbox from a real human with a real profile. The replies looked like this (paraphrased from the campaign):
Warm. Receptive. Same words, different channel.
A cold email arrives from a stranger at a domain nobody's heard of. A LinkedIn DM arrives from someone who just connected, whose photo you've seen, whose profile you clicked through, whose last three posts are sitting in your feed. The trust work is done before the DM opens.
That's why I say content is a role player and DMs score. The profile and the posts don't book the call. They make the DM land.
When you run cold email at scale, you risk your brand's reputation. That 12.5% opt-out rate is 42 people who now associate the company's name with something they had to unsubscribe from. Multiply that across a year of sequences.
Nobody blocks you for a thoughtful LinkedIn DM. Worst case, they don't reply. And no reply doesn't always mean no.
Here's the formula:
The hard part is No. 3. Here's how the DM works.
Open the conversation a few days after someone accepts your request. Don't send one multi-paragraph message. Send a sequence of short DMs a few seconds apart — something no automation tool can replicate, and 4-5 unread messages get opened.
The cold email sequence used these same beats. The difference is that a LinkedIn inbox lets you deliver them the way a human actually talks.

I hear this one a lot. Yes, you can send 40,000 AI-written emails into the void. You'll get 40,000 chances to make a bad first impression and a handful of "not interested" replies for your trouble.
Scale isn't the goal. Sales calls are the goal. 139 hand-written DMs beat 335 delivered emails by 22 calls to zero. If you're worried about scale, here's how I send 200 requests and 100 DMs a week without any automation.
Take Action: Pause your cold email sequence for two weeks. Take the exact messaging from it and rewrite it as the 5-message DM sequence above. Send it to 10 people who accepted a connection request from you this week. Compare the replies to your last 10 email replies. You'll have your answer before the two weeks are up.
Is LinkedIn outreach better than cold email for B2B? In my side-by-side test for an AI startup, yes, and it wasn't close: 335 emails delivered booked zero calls, while 139 LinkedIn DMs booked 22. The DM lands in a context where the prospect can see your face, your profile, and your posts. The email lands next to 40 other cold emails.
What is a good reply rate for LinkedIn DMs? 25-35% for new connections when the profile is optimized and the content is consistent. Profile viewers and post engagers convert at roughly double that. Cold email in my test replied at 5.7%, and every reply was negative or automated.
Can I run cold email and LinkedIn together? You can, and that's exactly what this client asked for. The email channel added cost, a 12.5% opt-out rate, and zero calls. If you have the budget for both, put it toward more DMs and better sales assets instead.
I put the full DM playbook into the 7-Step LinkedIn DM Guide: the sequences for new connections, profile viewers, and engagers, the follow-ups that revive dead conversations, and the pitch-slap hall of shame so you know what not to send. Free.
Or skip the learning curve. I run DM campaigns from executive accounts for clients every week — no automation, no cold email, 3-4 qualified calls a week.
Let's look at your ICP and map the first 10 DMs together. 30 minutes.