Standard B2B social-selling motion

Your LinkedIn DMs get ignored

Test Two LinkedIn DM Hooks

Send 50 LinkedIn DMs with hook A and 50 with hook B. Compare reply rate. Find what makes your ICP actually respond.

AcquisitionBeginner

2 weeks

When to use

Your ICP is active on LinkedIn (B2B, knowledge workers). You have a profile that doesn't look spammy.

If we DM 100 ICP-fit prospects on LinkedIn split between hook A (case-study) and hook B (question), one variant will get 2x+ the response rate.

The problem: LinkedIn DMs feel like spam — to senders and receivers. But the ones that work get 20–40% reply rates. Difference is the hook.

What works: Connect first (no note, or a one-line note). Wait 1–2 days. Then DM. Two variants:

A — Case-study hook: "Hey [name], saw you're leading [function] at [company]. We just helped [similar company] [specific outcome]. Open to a quick comparison?"
B — Question hook: "Hey [name], I'm researching how [persona] handle [problem] right now. You're one of the people I'd most want to learn from. Open to a 15-min call? Not selling."

How to do it:

Build a list of 100 ICP-fit LinkedIn profiles. Split randomly into two groups of 50.
Send connection requests on day 1.
On day 2–3, DM accepted connections with their assigned hook.
Follow up once after 4 days if no reply.
Track: connection accept rate, reply rate, positive reply rate, meeting booked.

Output: Winning hook + connection accept benchmark for your ICP. Reuse the winner; iterate the loser.

Source: Standard B2B social-selling motion.

7Impact

7Confidence

8Ease

gtmoutboundlinkedinab-test

Method

The problem: LinkedIn DMs feel like spam — to senders and receivers. But the ones that work get 20–40% reply rates. Difference is the hook.

What works: Connect first (no note, or a one-line note). Wait 1–2 days. Then DM. Two variants:

A — Case-study hook: "Hey [name], saw you're leading [function] at [company]. We just helped [similar company] [specific outcome]. Open to a quick comparison?"
B — Question hook: "Hey [name], I'm researching how [persona] handle [problem] right now. You're one of the people I'd most want to learn from. Open to a 15-min call? Not selling."

How to do it:

Build a list of 100 ICP-fit LinkedIn profiles. Split randomly into two groups of 50.
Send connection requests on day 1.
On day 2–3, DM accepted connections with their assigned hook.
Follow up once after 4 days if no reply.
Track: connection accept rate, reply rate, positive reply rate, meeting booked.

Output: Winning hook + connection accept benchmark for your ICP. Reuse the winner; iterate the loser.