We pulled sequence-level data from millions of B2B cold emails to find the right cadence length, channel mix, and inter-touch spacing. The answer: 6 touches over 14 to 18 days, split across email, LinkedIn, and a signal-based phone call.
The reply-rate curve by touch position
Where do replies come from across the sequence? The data:
| Touch position | Share of replies | Cumulative |
|---|---|---|
| Email 1 | 22% | 22% |
| Email 2 | 18% | 40% |
| LinkedIn connect | 9% | 49% |
| Email 3 | 21% | 70% |
| LinkedIn DM | 7% | 77% |
| Phone call (signal-based) | 11% | 88% |
| Email 4 (breakup) | 12% | 100% |
Three findings worth highlighting:
Finding 1: The first email produces only 22 percent of replies. Teams that send a single email collect less than a quarter of available pipeline.
Finding 2: Email 3 is the highest-yielding email in the sequence (21 percent of replies). This is the proof-point email; it tends to be where prospects who were warming convert.
Finding 3: The breakup email (final email) outperforms emails 2 and 3 individually, despite being last in the sequence. The "closing out" framing gives prospects a low-friction yes/no decision.
Reply rate by total sequence length
Cumulative reply rate as sequence length increases:
| Total touches in sequence | Cumulative reply rate | Marginal lift from previous |
|---|---|---|
| 1 | 1.4% | — |
| 2 | 2.9% | +1.5pp |
| 3 | 4.1% | +1.2pp |
| 4 | 5.0% | +0.9pp |
| 5 | 5.5% | +0.5pp |
| 6 | 5.9% | +0.4pp |
| 7 | 5.9% | +0.0pp |
| 8 | 5.7% | -0.2pp |
| 9+ | 5.4% | -0.3pp |
The curve is sharp at the beginning, flat at touch 6, and inverts at touch 8. Sequences longer than 6 touches do not produce more replies, and sequences longer than 8 produce fewer because spam complaints rise.
The channel mix effect
Holding sequence length constant at 6 touches, the channel mix changes the outcome substantially:
| Mix | Reply rate | Meeting rate |
|---|---|---|
| 6 emails only | 3.6% | 0.9% |
| 5 emails + 1 LinkedIn | 4.8% | 1.3% |
| 4 emails + 2 LinkedIn | 5.4% | 1.6% |
| 4 emails + 1 LinkedIn + 1 phone (signal) | 6.4% | 2.1% |
| 4 emails + 2 LinkedIn + 1 phone (signal) | 5.9% | 1.9% |
The 4-email + 1-LinkedIn + 1-phone mix wins on both reply rate and meeting rate. The additional LinkedIn touch in the 5-touch variant does not pay off in this slot; replacing an email with a phone call is the higher-leverage move.
Inter-touch spacing
The gap between touches matters. Reply rate by average gap:
- 1 day between touches: 4.1 percent reply rate. Too aggressive; reads as automation.
- 2 days: 5.2 percent. Working.
- 3 days: 5.8 percent. Optimal.
- 4 days: 5.6 percent. Working.
- 5 days: 5.0 percent. Slightly long; prospect forgets.
- 7+ days: 4.2 percent. Too long; the sequence loses continuity.
The 2-to-4-day spacing range is the working zone. Inside that range, 3 days is the sweet spot.
Total sequence duration
Inter-touch spacing combined with touch count produces a total duration. The data:
- Under 7 days: 4.2 percent reply rate. Too compressed.
- 7 to 14 days: 5.6 percent. Working.
- 14 to 21 days: 6.1 percent. Optimal.
- 21 to 30 days: 5.4 percent. Slightly long.
- 30+ days: 4.7 percent. Too long; the sequence loses thematic continuity.
The 14-to-21-day window is the sweet spot. Shorter sequences feel pushy; longer sequences feel forgotten.
Time-to-reply within the sequence
When during the sequence do replies actually arrive? The data:
- 21 percent of replies arrive within 24 hours of a touch.
- 43 percent within 72 hours.
- 78 percent within 7 days.
- 95 percent within 14 days.
- 5 percent arrive after the sequence completes.
The implication for reply triage: an active sequence produces most of its replies within a week of each touch. The reply backlog can be triaged daily without losing material pipeline.
The breakup email effect
The final email in the sequence is the second-highest-reply email after email 3. The breakup framings that work:
| Format | Reply rate (this email only) |
|---|---|
| "Should I close the loop on this?" | 6.8% |
| "Closing out unless you tell me otherwise" | 6.2% |
| "Mind if I close the loop?" | 5.9% |
| "Last note from me on this" | 5.1% |
| "Final follow-up" | 4.4% |
| "Did not hear back, let me know" | 3.2% |
The "close the loop" framing wins consistently. The phrasing implies finality without finality. It gives the prospect a low-friction yes/no decision and triggers more "actually, let me see what you have" replies than any other format.
What does not work
Several common patterns underperformed in our data:
- "Bumping this up" or "in case you missed it": Reduce reply rate by 1 to 1.5 percentage points on the touch they appear on. Reads as automation.
- Forwarding previous emails in a thread: Long threaded chains reduce reply rate by 0.8 to 1.2 points after the third touch. Start a new thread for emails 4 onward.
- Multi-day touches on the same day: Two emails on the same day reduce reply rate on the second touch by 60 to 70 percent. Spread.
- Re-engaging silent prospects after 30 days: A second sequence to prospects who completed the first sequence without replying produces only 1.1 percent reply rate (vs 5+ percent on fresh prospects). Move silent prospects to a 90-to-120-day nurture cycle instead.
Five sequence takeaways
- 6 touches over 14 to 18 days is the sweet spot for most B2B outbound.
- 4 emails + 1 LinkedIn + 1 signal-based phone call is the highest-yielding channel mix.
- 3-day spacing between touches is optimal. 2 to 4 days is the working zone.
- The breakup email is the second-highest reply driver after email 3. Use it; phrase it as "close the loop."
- Do not extend beyond 7 touches. Reply rate plateaus and spam complaints rise.