Cold email follow up timing: what 90 days of sequence gap testing showed about optimal days between emails
sequence_gaps_tomasz · 2026-09-29 · 0 views
I spent 90 days testing follow up timing variations on the same ICP segment. Same copy. Same list quality. Different gaps between emails. Here is what the data actually showed.
The test setup. Four sequence variations across a matched ICP. 300 contacts per variation. All lists run through ZeroBounce before loading. All inboxes warmed and provisioned through PuzzleInbox. Sequences ran in Instantly. The only variable was the number of days between each email.
Variation A: 2 days between each email. Reply rate: 2.9%. High complaint signals. Several contacts replied asking me to stop before they reached the third email. Too aggressive. You feel pushy before you have established any trust.
Variation B: 4 days between each email. Reply rate: 3.8%. Fewer aggressive negative replies. Most positive responses came on email 2 or 3. Felt closer to a natural pace.
Variation C: 7 days between each email. Reply rate: 4.1%. Small improvement over 4 days but the sequence stretched over three weeks total. For a 4-email sequence that means the last touch lands almost a month after first contact. Works fine for enterprise or long consideration cycles. Too slow for SMB where attention spans on cold outreach are shorter.
Variation D: Mixed cadence of 3, 5, 7 days. Reply rate: 4.4%. Best performer across all four. Short first gap while your first email is still fresh in their memory. Then progressively more space as the sequence continues. Felt natural. Fewer complaint signals than the flat 2-day approach by a wide margin.
What I run now. Default cadence of 3, 5, 7 days for all sequences. The first follow up comes while the context is still fresh. The later follow ups give the prospect room without losing the thread of the original email. Adjust for ICP deal cycle length but start there and only deviate when you have data that tells you to.
Flat spacing is a default, not a strategy. Test your timing the same way you test your copy. It moves reply rates more than most people expect.