Lifecycle and segmentation

Sending everything to everyone is the most common cause of email decline, and it is invisible until it is a deliverability problem.

Why it damages you

Every send to someone uninterested produces a small negative signal — no open, no click, sometimes a complaint. Do it repeatedly and mailbox providers learn that your mail is ignored, which affects delivery to the people who do want it.

Segmentation is not a sophistication play. It is protection.

The welcome sequence

The highest-engagement moment you will ever have with a subscriber is immediately after they sign up. A welcome sequence uses it: confirm what they will get, deliver the thing promised, and give one clear next step.

It also serves a technical purpose — early positive engagement establishes reputation with that recipient's provider before you start regular sending.

Engagement segments

Segment by recency of engagement, not by demographics:

• Active — opened or clicked recently. Send normally.
• Lapsing — nothing in a couple of months. Reduce frequency, change approach.
• Dormant — nothing in six months or more. Stop, or run one re-engagement attempt.

The precise windows depend on your cadence. The principle does not.

Sunsetting

Removing people who have not engaged for a long time. It feels like deleting an asset, and it improves results — your metrics rise because they were being dragged down, and your reputation improves because you stopped sending to people who ignore you.

The recycled-spam-trap risk makes it more than a hygiene preference. A long-dormant address may no longer belong to a person at all.

Run one clear re-engagement email — "do you still want these?" — then remove.

Frequency

There is no universal right answer; there is a promise you made at signup. Consistency beats optimisation, and a sudden increase in frequency is one of the most reliable ways to raise complaints.

Open rate is broken as a segmentation input

Apple Mail Privacy Protection pre-fetches images, marking messages as opened whether or not anyone looked. A meaningful share of your "opens" are fictional.

Consequence: segment on clicks, not opens, wherever the decision matters. Opens remain weakly useful in aggregate for trend, and useless for individual targeting.

What trips people up

• Sending the whole list every time because the list is the number that gets reported.
• Segmenting on opens post-MPP.
• Never sunsetting, then wondering why Gmail placement fell.

You have got this when

You can say who you did not send the last campaign to, and why.