Recap
- Send three to five touches total. The first three carry most of the replies. Past five you mostly burn goodwill and sender reputation.
- Space them tight early, loose late: roughly day 0, day 3, day 7, day 14, day 24.
- Every follow-up has to add something. A new angle, a proof point, a different question. Never send the word bump.
- Stop on a reply, a hard bounce, a clear no, or your last planned touch. Honor every opt-out the moment it lands.
The right cold email follow-up sequence is three to five total touches, spaced roughly three to seven days apart, with the gaps widening as you go. The first three messages do most of the work. Each follow-up should carry a new reason for the prospect to care, not just a reminder that you exist. Send them from an authenticated inbox, stop the moment someone replies or bounces, and never chase past five.
How many follow-ups should you actually send?
Three to five touches, counting the opener. That range covers almost every reply you were ever going to get. Plenty of people answer on the second or third nudge, not the first, because the first one landed on a bad day. But response rates fall off a cliff after the third or fourth, and a sixth or seventh email mostly trains people to mark you as spam. That hurts every future send from your domain.
Follow-ups are not about wearing someone down. They are about catching a busy person on the one day your timing fits their problem. A few well-spaced, useful touches do that. A relentless drip does not.
How far apart should the touches be?
Tight at the start, looser toward the end. The first follow-up can come three to four days after the opener, while you are still faintly in memory. Then stretch: five days, then seven, then ten or more before the final touch. The widening gap respects that someone who has not replied by touch four is not going to be moved by you emailing faster. It also keeps you from hammering the same person inside a single week, which is the fastest way to get reported.
One practical note. Skip weekends and obvious holidays for the first couple of touches. A Tuesday or Wednesday morning beats a Friday afternoon for getting actually read.
Why does every follow-up need to add something?
Because a follow-up that only says checking in or just bumping this gives the reader nothing new to react to, so they do the easy thing and ignore it again. Each touch should change the offer. Lead with a different angle. Add a proof point. Ask a smaller, easier question. Share something genuinely useful with no ask attached. You are giving the person a fresh reason to reply, not reminding them of the reason that already did not work.
A simple way to plan it: assign each touch a single job.
- Touch 1. The core pitch. One clear reason you are reaching out and one specific ask.
- Touch 2. A different angle on the same problem, or a short proof point that you understand their world.
- Touch 3. Lower the bar. Ask a yes or no question, or offer something useful with no meeting attached.
- Touch 4. A pattern or insight from similar companies, kept honest and specific.
- Touch 5. The polite close. Name that you will stop, and leave the door open.
What does a simple cadence look like?
Here is a five-touch sequence you can run as a default and adjust by segment. Days are counted from the first send.
| Touch | Day | Job | Channel |
|---|---|---|---|
| 1 | 0 | Core pitch, one ask | |
| 2 | 3 | New angle or proof point | |
| 3 | 7 | Easier question, or give something useful | Email plus optional LinkedIn view |
| 4 | 14 | Insight from similar companies | Email or LinkedIn message |
| 5 | 24 | Polite close, door left open |
Three touches is a fine sequence on its own if you would rather stay light. Just compress it: day 0, day 4, day 10. The shape matters more than the exact dates.
How should you mix in other channels?
Email carries the sequence, and other channels support it. A LinkedIn connection or a comment on something the prospect posted, timed near touch three or four, makes your name familiar so the next email reads as warmer. A short, well-aimed call can work for higher-value accounts. The mistake is treating every channel as a separate full-blown sequence. That is not multichannel, that is the same person getting pestered four ways at once, and it gets you blocked.
Keep one source of truth for whether a person has replied. If they answer on LinkedIn, the email sequence stops. If they answer email, the calling stops. The channels should act like one coordinated effort, not four teams who never talk.
How do you handle replies and know when to stop?
The instant someone replies, the automated sequence stops and a human takes over. That is the whole point of running a sequence: get to a real conversation, then drop the script. Sort replies into interested, not now, and not a fit, and answer the live ones fast. Speed matters more than polish here.
Stop the sequence on any of these: a reply, a hard bounce, a clear no, or your last planned touch. Honor unsubscribe and stop-emailing-me requests immediately, every time, no exceptions. The 2024 bulk sender requirements from Google and Yahoo made one-click unsubscribe and low complaint rates non-negotiable, and ignoring an opt-out is the quickest way to torch your domain. A silent prospect, though, is not a permanent no. Park them and come back in a quarter with a genuinely new reason, like a product change or something that shifted in their company.
Running this by hand across hundreds of prospects is where it falls apart. People forget to stop a sequence after a reply, send touch four to someone who already said no, or let the gaps drift. This is the kind of disciplined, stateful bookkeeping that software handles better than a human with a spreadsheet. A tool like LaunchSurface keeps the cadence, the channel coordination, and the stop conditions straight so the sequence stays polite at scale. The principles above hold whether a person or a system is running them.