Sequence length

The right number of emails in a sequence depends on deal size, cycle length, and your patience. Most B2B teams default to 5-7 touches. Below 3 and you leave money on the table. Above 10 and you annoy your ICP.

The standard by context

SMB (under $3K ACV, short cycles)

3-5 emails over 10-14 days. Impulse buyers decide fast. Long sequences waste tokens.

Mid-market ($3K-$30K ACV)

5-7 emails over 21-28 days. Standard default.

Enterprise ($30K+, long cycles)

7-12 emails over 6-10 weeks. Multiple stakeholders, longer decision windows.

High-touch partnerships

3-5 emails over 10-14 days, then pause. Re-engage 3-6 months later with new angle. Repeat.

Reply rate by email position (B2B typical)

Aggregate 5-touch reply rate: 8-15%. Most teams stop adding touches after 5-7 because the yield drops sharply.

The diminishing returns curve

Adding email 6 to a 5-touch sequence adds about 0.5-1% more reply rate. Email 7: 0.2-0.5%. Email 8+: marginal.

But each additional email burns more sender reputation, more list goodwill, more annoyance risk. The ROI per touch drops dramatically after email 5-6.

When longer sequences make sense

When shorter sequences make sense

The re-engagement strategy

After the 5-touch sequence ends, don't delete non-responders. Wait 3-6 months. Re-engage with:

Many deals close on re-engagement 8-12 months after first outreach.

The "never-ending sequence" anti-pattern

Some operators set up sequences that never end, automated bumps every 2 weeks forever. Don't.

A clean 5-touch sequence, clean breakup, clean re-engagement 6 months later beats perpetual nagging every time.

Testing sequence length

A/B testing sequence length
Split 1000 prospects in half:
- Group A: 5-touch sequence
- Group B: 7-touch sequence

Measure:
- Total reply rate
- Positive reply rate
- Meetings booked
- Spam complaints
- Unsubscribe rate

Run 4-6 weeks. Calculate per-email marginal contribution. Pick the length where adding another email adds less than 0.5% reply rate and increases complaints or unsubscribes.