How we get 92% open rates on LinkedIn DMs

If you don’t know me already, my name’s Jack C.

I don’t do bro-marketing.
I don’t sell a dream.
And I definitely don’t promise “instant inbound leads” from a profile banner.

But what I do have is over 4 years of experience sending LinkedIn DMs that actually work.

We’ve tested over 10,000 cold messages.
Across multiple offers, price points, niches, and industries.

Here’s the strategy that now gets us:

→ 89% connection acceptance
→ 92% open rate
→ 42% response rate

And no, it’s not about sending “Just following up” 6 times in a row.

Here’s the full breakdown 👇

1. Send connection requests without notes
Notes tank your acceptance rate.
They immediately signal: “I’m here to sell you something.”
We tested this at scale — note-free requests got accepted 38% more often.

2. Personalise the first message with context
When someone accepts, don’t jump straight to your offer.
Instead, reference something that shows you paid attention.
e.g.
→ "Noticed you’re scaling the sales team"
→ "Saw your post about outbound"
→ "Loved what you’re doing with [specific thing]"

3. Ask ONE low-friction question
Something they can answer in 5 seconds.
e.g. “How are you handling [problem] right now?”
No long pitch. No intro paragraph. Just signal + curiosity.

4. Follow up with a voice note
Less than 0.001% of LinkedIn messages are voice notes.
They cut through the noise instantly and humanise your message.

That’s it.

No fancy funnels.
No bait-and-switch tactics.
Just a simple framework that starts conversations, not sales pitches.

And conversations → calls → clients.

Watch below as The chief, Alfie, breaks down the whole strategy A–Z — with examples, templates, and benchmarks.

— Jack C. @ Team Prosp 👊

P.S Struggling to consistently book calls from LinkedIn?

If so, try Prosp – the only LinkedIn automation platform that

• Sends AI personalized DMs

• Sends custom voice notes whilst you sleep

• Finds leads from LinkedIn search, sales nav, events, groups and post engagement.

• Turns content → booked calls