Every year, somewhere around September, the same idea surfaces in a planning meeting. Peak is coming, the targets are up, and there are 400,000 people on the database who have not heard from us in a while. Why not wake them up before the big push?
It is the single worst piece of timing in email marketing, and it is proposed with complete confidence in boardrooms every autumn.
Re-engagement is not a campaign. It is a deliberate decision to send mail to the people least likely to respond, at a moment when how people respond is the thing deciding whether your email arrives. Done at the wrong time it damages the exact weeks you were trying to protect.
Which does not make it wrong. It makes it a risk decision, and risk decisions need assessing rather than scheduling.
Why re-engagement is a deliverability event rather than a campaign.
The peak season rule, and the maths behind it.
Why disengaged data is not one group, and the high, medium and low risk split.
Why the medium tier is a stakeholder conversation rather than a technical one.
How to run one properly when the timing is right.
And what to do with the people who do not come back.
Mailbox providers decide where to put your mail based largely on how people respond to you, assessed as a rolling history rather than a snapshot. Every send contributes evidence.
A re-engagement campaign is, by definition, a send to the population you already know does not respond. You are volunteering a large block of negative evidence, deliberately, in one go.
The signal you send is the wrong one, at volume. Low opens, low clicks, deletions without reading. Providers read that as a sender whose mail is unwanted, and they apply the conclusion to everything you send afterwards.
Complaints spike, because people do not remember you. Somebody who has not heard from you in two years does not think I have lapsed. They think who is this, and the report button is closer than the unsubscribe link.
Dormant data is where recycled spam traps live. An address abandoned long enough gets reclaimed by the provider and repurposed as a trap. Your disengaged segment is, almost by definition, the oldest and least maintained part of your list, which is exactly where they accumulate. Re-engagement sends are one of the most common ways businesses hit them.
And the damage is not confined to the campaign. Reputation is shared across whatever sends from that domain and infrastructure, including the transactional mail you cannot afford to lose.
My position on this does not have exceptions, which is unusual for me. Do not re-engage disengaged data in the run-up to peak season, a product launch, a fundraising appeal, or any send the year depends on.
Reputation is a rolling assessment, not a daily reset. A poor send in early October is still part of the picture providers are using in November. You do not get a clean slate because the campaign ended.
Recovery runs in weeks, and the gap you left is shorter than that. Most businesses schedule the reactivation four to six weeks before peak, which is comfortably inside the window where the damage is still being counted.
Peak is when you are sending at your highest volume anyway. Adding a reputation problem to your busiest sending period is how a difficult quarter becomes a lost one.
And the upside is small, by definition. Reactivation rates on long dormant data are low. You are risking placement for your whole active audience to recover a thin slice of an inactive one.
In the quiet period. After peak, well before the next one, when a few weeks of impaired placement costs you least and you have time to recover properly before anything important goes out.
For most businesses that means January or February rather than October, which is the opposite of when it usually gets proposed.
The conversation nearly always starts in the wrong place. Somebody asks whether we should re-engage the lapsed contacts, as though the 400,000 people in that segment are a single population with a single answer.
They are not. Some of them are a clear and present danger to your sending reputation. Some of them are harmless to include. Most sit in between, and the decision about them is not a technical one at all.
So the first job is not writing the campaign. It is splitting the disengaged population into risk tiers and deciding what happens to each one.
Marketers want the medium tier resolved with a threshold, because a threshold can be defended in a meeting. The reason it cannot be is that the trade-off is commercial rather than technical.
Including the medium tier buys a slightly larger reactivation pool at the cost of some reputation risk. Whether that is a good trade depends on how much placement your business can afford to lose, what is scheduled in the following six weeks, and who carries the consequences if it goes badly.
None of which is yours to decide alone, and none of which should be decided by whoever is loudest in the planning meeting.
Present it as a trade, not a recommendation. Including this group gives us roughly this many more contacts and this much more risk, at this point in the calendar.
Put the downside in business terms. Not complaint rates and reputation. Lost orders during peak, and the password resets that stop arriving if it goes far enough.
Name who signs it off. Somebody accountable for the outcome, in writing, not a nod in a meeting nobody minuted.
Agree what would make you stop mid-campaign. A complaint rate, a placement change, a bounce threshold. Decided in advance, when nobody is invested in continuing.
And hold out a control group so you can prove what happened. The argument recurs every year otherwise, from memory, with both sides remembering it differently.
Send in waves, smallest risk first. A single send to the whole disengaged population is the volume spike to a cold audience that causes the damage. Split it across weeks and watch between waves.
Re-verify before you send. A verification pass on data that has sat untouched for a year removes the dead addresses before they can hard bounce, and some of them before they become traps.
Lead with relief, not guilt. The most effective re-engagement email I know offers people less of you, not more. Would you rather hear from us monthly? Or not at all, with one click? Nobody has ever been won back by being asked whether they still love you.
Make it useful on its own terms. Somebody who ignores the email should still have got something from the subject line and first line. You are rebuilding a reason to be in their inbox, not running a promotion.
Keep it short. Two or three emails, not a programme. If three well-made attempts do not move somebody, a fourth will produce a complaint rather than a conversion.
Watch complaints daily while it runs. Weekly reporting on a three week campaign tells you about the problem after the campaign has finished causing it.
And stop when you said you would stop. The threshold you agreed before you started, honoured without renegotiation.
The point of a re-engagement campaign is not only reactivation. It is earning the right to suppress everybody who did not respond, with evidence, so the argument does not happen again next year.
Suppress, do not delete. Keeping the record means somebody returning through another route is recognised rather than treated as new, and it keeps your consent history intact.
Make the suppression span every system. A contact removed from the sending platform but still live in the CRM comes back on the next sync, and you have achieved nothing.
Report what it cost and what it returned. Reactivated contacts, complaints, unsubscribes, bounces, and any placement movement. On one page, while people still remember agreeing to it.
And set the cadence so it never builds up again. A standing sunset process means you are never again looking at a 400,000 contact decision. The annual crisis is a symptom of not having one.
Everything above is the principle. The part that takes longer is the assessment itself: what puts a contact in the high, medium or low tier, how to build the segmentation so it is repeatable rather than a one-off judgement call, and how to run the stakeholder conversation with evidence in hand.
That is the Re-engagement Email and Disengaged Data course, and it covers the framework end to end.
It is also included free as a bonus course with the Email Deliverability Certification, which is where I would point you if re-engagement is a symptom rather than your whole problem. Most businesses asking whether they should reactivate their dormant data have a wider set of questions underneath it, about list quality, sending reputation and what is reaching the inbox in the first place.
The Certification. CPD certified, with live sessions, and you run a full audit on your own email marketing as you go. You finish knowing exactly where your mail is landing and what to fix in what order.
The re-engagement course, included. The risk framework, the segmentation and the execution, at no extra cost as part of the programme.
Re-engagement gets treated as a campaign because it looks like one. It has a creative, a send date and a target, so it goes in the calendar alongside everything else.
What it is, underneath, is a decision to spend sending reputation on the least responsive part of your audience. Sometimes worth it. Never worth it before peak. And never worth it to the whole disengaged population at once, because they were never one population.
Split them by risk, exclude the high tier, have the real conversation about the middle, and run it in January when the cost of getting it wrong is lowest.