Between 90 and 97% of the people in a finished cold email sequence never reply. Most teams write that list off and buy more leads. That is backwards: every one of those people already passed your ICP, was verified deliverable, and cost money to find.

Going back to them is the cheapest qualified reach you have, and also the easiest place to damage a sending domain. Go back too soon and it reads as pestering. Send to a bounced address twice or to someone who asked you to stop, and the complaints land on your domain. Resend the same sequence with a new subject line and you get the same result with worse odds, because they already ignored that exact argument once.

What it does

The prompt walks you through five steps, one at a time, waiting for your answer before moving on: check the timing, pull the eligible set, re-verify, choose the new reason, and set the shape of the sequence. At the end you get a plain-text plan with the counts, the exclusions, the new first line, the volume, the thresholds to watch, and the date of the next pass. If a platform is connected, it pulls the lists itself and asks before anything that costs credits or changes a live campaign.

What it catches

• Timing windows. Under 60 days, wait. Sixty to 90 days is the window. Up to a year is fine and often better. Over a year, treat it as a fresh list.

• Three permanent exclusions. Anyone who said no or unsubscribed, anyone who bounced, and anyone in another live sequence. A platform's unsubscribe handling may only cover its own sequences, so keep your own suppression list too.

• The re-enrollment block. Most sending platforms block re-enrolling someone who finished a sequence. Lift it knowingly for a list you built on purpose, never to make a number go up.

• Data decay. Over 60 to 90 days people change jobs and addresses die, so the list gets verified again before anything sends.

• Job changes are leads, not decay. Someone who moved company is a new prospect with a warm prior touch, and goes into a fresh campaign built on that move.

• A real reason in the first line. A new signal, new proof, a different angle, a different person at the same company, or honest acknowledgement. "Just following up" is not a reason.

• Two steps and lower volume. They already had a full sequence, and a recipient who remembers you is more likely to complain, so bounce and complaint rates get watched harder.

• Separate tracking. Re-engagement reply rates are not comparable to cold ones, and mixing them corrupts both numbers.

How to use it

Pick whichever route fits your setup.

1. Paste it into a chat. Copy the prompt below into ChatGPT, Claude, or any other assistant and tell it which sequence finished and when.

2. Add it to a Claude Project. Upload it as a Project file. Connect Apollo and it pulls the list and runs the exclusions itself: the Claude setup guide.

The prompt

Copy the whole prompt below, from the first line to the last.

You are helping me go back to the people who never replied to a finished cold email sequence. Somewhere between 90 and 97% of a sequence never replies, and most people treat that as a dead list and buy more leads. That is backwards. Every one of them already passed my targeting, was verified deliverable, and cost money to find. They are the cheapest qualified reach I have.

Work through the steps below with me one at a time. Ask, wait for my answer, then move on.

If a sending platform or prospecting database is connected as a tool, pull the lists and run the checks yourself, and confirm with me before anything that costs credits or changes live state: re-verifying, enrolling, activating. If nothing is connected, tell me exactly what to pull and I will bring it back.

You advise and I decide. Timing, exclusions, and volume are the exceptions: those protect my sending domains, so be firm about them.

The stance. Silence is not rejection. A non-reply means the message did not land at that moment, in that inbox, in that framing. It does not mean a bad fit. A list built in March is not worthless in September; it needs a new reason to hear from me. What re-engagement is not: the same sequence again with a new subject line. The recipient has already ignored that exact argument once, so it is the same attempt with worse odds.

Step 1. Check it is safe to go back. Ask when the sequence finished.

Since it finished	Do
Under 60 days	Wait. Too soon reads as pestering and earns spam complaints, which cost the domain.
60 to 90 days	The window. Long enough to be forgotten, short enough that their situation is recognisable.
90 days to a year	Fine, often better, because something has changed. Expect more job changes and stale data.
Over a year	Treat it as a fresh list. Re-verify everything.

Step 2. Pull the eligible set, minus three permanent exclusions. Remove, no matter how much time has passed:

• Anyone who said not interested, asked to stop, or unsubscribed. Permanent, not a timing problem. A platform's own unsubscribe handling may only cover sequences sent from that platform, so these people belong in my own suppression list too.

• Anyone who bounced. Sending to a bad address a second time is how a domain dies twice.

• Anyone in another live sequence. Check before enrolling, not after.

Most sending platforms block re-enrolling someone who finished a previous sequence. That block is the right default, and lifting it is the whole decision. Lift it knowingly, for a list we deliberately assembled, never to make a number go up. Switching it on quietly is how a list gets burned.

Step 3. Re-verify before spending or sending. Data decays fastest in the fields that matter. Over 60 to 90 days, people change jobs and addresses go dead. Run the list through verification again and drop anything no longer deliverable. That is far cheaper than a bounce spike.

A job change is an opportunity, not decay. Someone who moved company is a new prospect with a warm prior touch, and someone new in a role has budget and a mandate to change things. Route them into a fresh campaign built on that move, not into the re-engagement sequence.

Step 4. Choose the new reason. This decides whether it works. Something must be different, and they have to see it in the first line. Help me pick one:

• A new signal. They are hiring, they raised, they adopted a tool, they visited my pricing page. The strongest option, because it is about them, right now.

• New proof. A result or case study I did not have last time, strongest when it is in their industry.

• A different angle on the same offer. Led with cost last time; lead with speed or risk now.

• A different person at the same company. A second route into the account, not a second attempt at the person. Respect the cap on contacts per company.

• Honest acknowledgement. "I wrote to you in March about X, it clearly was not the moment" is disarming and true. Use it sparingly, never as a guilt trip.

"Just following up" and "bumping this to the top of your inbox" are not reasons. They admit there is nothing new to say, and they read that way.

Step 5. Keep it shorter and slower.

• Two steps, not three. They have already had a full sequence, and total volume received decides whether I read as persistent or as a problem. If two touches with a new angle do not land, stop and put them back in the pool.

• Lower daily volume than a cold campaign. The recipient may remember me, so complaint risk per send is higher. Watch bounce and complaint rates harder than usual, and know how to stop the campaign immediately before it starts.

Cadence and tracking. Make it a standing quarterly pass: everything that finished more than 60 days ago, minus the exclusions, gets one new angle, so the pool never silently rots. Track it separately from cold campaigns. Re-engagement reply rates are not comparable to cold ones, and mixing them corrupts both numbers.

What to hand me at the end: a plain-text plan with the source sequence and when it finished, the eligible count and how many each exclusion removed, the re-verification result and the job changes routed elsewhere, the new reason and the first line it implies, the two-step shape, the daily volume and the bounce and complaint thresholds to watch, and the date of the next quarterly pass.

Method adapted from the re-engagement skill in Apollo Operator, a free, open-source headless GTM toolkit by Creatop: github.com/creatop-gtm/apollo-operator

Part of the Apollo Operator prompt pack

This is one of 19 free prompts from Apollo Operator, the open-source headless GTM toolkit Creatop builds and runs on its own campaigns. Get the full pack here: the Apollo Operator prompt pack.