How many follow-ups should you send in a cold email sequence?

March 24, 2026

I tracked every reply across 14 cold email campaigns over the past year. 2,800 total sends for Scouter and Prospect Organic. Here’s the one number that changed how I think about follow-up:

67% of my replies came from touches 2–4. Not the first email.

Most people send 1 email, maybe 2, and stop. They’re quitting before the results start.

Reply rate by touch number

Here’s the breakdown across all 14 campaigns, normalized to a 200-prospect campaign with an 18% cumulative reply rate (36 replies):

Touch 1 gets the most replies per touch, sure. But touches 2–4 combined account for 67% of total replies. If you stop after touch 1, you’re leaving 24 replies on the table out of every 36.

Touch 3 is consistently the highest-performing follow-up. That’s not random. By day 7, the prospect has seen your name 3 times. They’ve had time to look you up. The sequence architecture matters – touch 3 is where you offer something tangible (a free resource, a case study, a custom list). That changes the dynamic from “someone asking for my time” to “someone giving me something useful.”

The case for exactly 4 touches

I’ve tested 3-touch, 4-touch, and 5-touch sequences. Here’s what the data says.

3 touches – Gets about 85% of the results of a 4-touch sequence. Viable if you’re sending high volume and want less complexity. But you lose the breakup email, which consistently pulls in replies from people who ignored everything else.

4 touches – The sweet spot. You get 95%+ of the total replies you’re going to get. The 4th touch (breakup) creates a natural ending that gives the prospect one last reason to respond. My 4-touch sequences average 18% cumulative reply rate across all campaigns.

5+ touches – Diminishing returns hit hard. Touch 5 in my tests added less than 1% to the cumulative reply rate. But it did something worse – it increased unsubscribe requests and “please remove me” replies by 3x compared to 4-touch sequences.

The math is simple: touch 5 costs you more goodwill than it earns you replies.

Here’s the data from a head-to-head test. Same list, same offer, same copy for touches 1–4. One group got a 5th touch on day 16. The other didn’t.

The 5-touch sequence got 1 more total reply – but 6 more negative replies. Net positive replies actually went down. That 5th email didn’t earn a conversation. It earned resentment.

Why touch 3 outperforms touch 2

This confused me at first. Touch 2 is closer in time to the original email. Why does touch 3 pull more replies?

Two reasons.

1. Touch 3 changes the approach. In a well-structured sequence, touch 2 is a gentle follow-up that adds a proof point. Touch 3 shifts entirely – it offers something for free. A custom list. A teardown. A relevant case study. That shift from asking to giving is what triggers the reply.

2. Familiarity builds by day 7. The prospect has seen your name in their inbox 3 times. You’re not a stranger anymore. You’re “that person who keeps reaching out about [topic].” When you then offer something useful, the barrier to replying drops significantly.

This only works if each touch adds new value. If your follow-ups are “just bumping this to the top of your inbox,” touch 3 won’t outperform anything. It’ll just annoy people. Read how to follow up without being annoying for the framework.

Why more than 4 hurts

Beyond the reply rate data, there are 2 structural reasons to stop at 4.

Deliverability. Email providers track engagement signals. If your 5th and 6th touches consistently get ignored or marked as spam, your sender reputation takes a hit. That affects deliverability on all your emails – not just the ones to that prospect. For solo founders running outbound from their primary domain, this is a real risk.

List economics. Every prospect who gets annoyed by a 5th touch is someone you can’t re-approach later. You’ve burned the bridge. With a 4-touch sequence, you end gracefully. That person is still approachable in 90 days with a new angle.

The 90-day re-engagement rule

A prospect who didn’t reply to your 4-touch sequence isn’t dead. They’re just not ready.

Wait 90 days. Then re-engage with a completely new angle.

Not “following up on my email from 3 months ago.” A fresh sequence. New subject line. New observation. New reason for reaching out.

Subject: [New angle about their company]

Hey [first name],

Different topic from my last outreach – noticed [company]
just [new development: product launch, funding round, new hire,
public initiative].

[New value prop or question relevant to the new development.]

– Joe

I run 90-day re-engagement on about 30% of my prospect lists. The results are telling:

• Re-engagement reply rate: 11% (vs. 6% on touch 1 of the original sequence)
• 40% of re-engagement replies reference the original sequence: “I remember your emails – timing is better now”
• 0 negative replies across 3 re-engagement campaigns

The original sequence did its job even though they didn’t reply. It planted your name. It established credibility. The 90-day gap removed the pressure. The new angle gave them a fresh reason to respond.

This only works if you ended the original sequence respectfully. If your 6th touch was “I guess you’re not interested,” there’s no coming back from that. End at 4. End gracefully. Keep the door open.

The system

Here’s the complete framework:

1. 4 touches. Days 0, 3, 7, 12. Each touch adds new value or shifts the angle.
2. Stop at 4. No exceptions. The data doesn’t support a 5th touch.
3. Wait 90 days. Mark non-responders for re-engagement.
4. Re-engage with a new angle. Fresh sequence, fresh observation, fresh reason.
5. Track per-touch metrics. If touch 3 isn’t your best follow-up, your sequence architecture needs work.

The biggest mistake in cold email isn’t sending too few messages. It’s sending too few good messages. 4 well-crafted touches will outperform 7 lazy ones every time.

Build your sequence with these examples. Write subject lines that earn the open with these patterns. And if you’re running multi-channel outbound, the same 4-touch principle applies – just spread across channels.