Email list hygiene: how to clean your contact list

Email list hygiene is the ongoing practice of removing the addresses that can't receive your mail, or shouldn't, so that the people who can will actually see it. Hard bounces, role-based aliases, spam traps, and long-term inactives all belong on the removal list. Good senders treat it as a deliverability workflow with a schedule, not a chore they get to when the list looks messy.

Lists decay on their own. People change jobs, abandon accounts, mistype at signup, and lose interest without telling you. Leave a list untouched for two years and a meaningful slice of it is dead weight dragging your bounce rate toward the ceiling Gmail enforces under its bulk-sender rules (Google Postmaster, 2024). Clean the list, or the filters will clean it for you, one spam-folder placement at a time.

What email list hygiene is

It's the repeatable process of identifying and removing addresses that either can't receive mail or shouldn't for deliverability reasons: hard bounces, spam complaints, role-based aliases, catch-all domains you can't verify, known spam traps, and subscribers who haven't engaged in months. Part data-quality work, part sender reputation management.

Think of your list the way a gym owner thinks of a member roster. The names on the wall don't matter; the ones who show up do. Every dead address on your file is a small vote against your next campaign, and they compound.

List hygiene is a deliverability problem

Mailbox providers score senders on engagement (opens, clicks, replies, deletes without reading, complaints) and on bounce rates. A list full of dead addresses drags all of those in the wrong direction simultaneously. You don't just lose the dead subscribers; you poison the channel for the live ones.

Google's Postmaster Tools documentation is explicit about the complaint side: rates above 0.30% trigger filtering, and sustained rates above 0.10% already degrade reputation. Gmail is generally the strictest of the big providers, though Yahoo, Microsoft, and Apple apply comparable thresholds.

The knock-on effect is what actually hurts. A 5% bounce rate on a 50,000-contact send puts 2,500 unknown users in front of ISP filters, and those filters don't conclude "this list is old." They conclude "this sender doesn't know who they're mailing," then start throttling the other 95%. Your engaged subscribers, the ones who open and click and buy, stop seeing you. That spiral is the thing list hygiene exists to prevent; watch inbox placement alongside bounces and you'll catch it early.

The five kinds of bad address

Bad addresses fall into five buckets, each with its own signal and its own fix. You find them through some mix of bounce logs, engagement data, syntax checks, and third-party verification.

A hard bounce is a permanent failure; the mailbox doesn't exist, the domain doesn't resolve, or the address was typed wrong. SMTP returns a 5.x.x code, and any reputable ESP auto-suppresses on the first one. Check that yours actually does; some older tools retry indefinitely, which is the single most damaging default in email software.

A soft bounce is temporary: mailbox full, server down, message too large, returning a 4.x.x code. One soft bounce is noise. Three consecutive soft bounces from the same address is a dead mailbox with a polite error message, and it should graduate to hard-bounce suppression.

Spam complaints come back through a feedback loop when a subscriber hits "Report spam" (Gmail is famously quiet about individual complaints, though the aggregate rate shows up in Postmaster Tools). Remove complainers immediately. Keep a running spam complaint rate and treat any upward drift as urgent rather than interesting.

Long-term inactives are contacts who haven't opened, clicked, or replied in six to twelve months. Either they're gone or they don't want your mail; either way they're engagement poison. The right window depends on cadence. A weekly newsletter should flag 90 days of silence as cooling, while quarterly B2B sends can reasonably stretch to twelve months.

Role-based addresses are shared aliases like info@, sales@, support@, and admin@. Nobody personally opted in to a shared mailbox, and they complain at higher rates than named inboxes. Catch-all domains are the related headache; they accept any address as valid even when the mailbox doesn't exist, so verification returns "unknown" rather than a useful answer.

Spam traps and how they reach your list

Spam traps are addresses run by blocklist operators, ISPs, and anti-abuse groups (Spamhaus, Abusix, SpamCop among them) to catch senders with careless list practices. Hitting one can land you on a public blocklist that every major receiving server checks. Three flavors exist, and their origin stories differ enough to matter.

Pristine traps were never subscribed to anything. Operators plant them on obscure web pages and in data that scrapers harvest, so the only route onto your list is scraping, either yours or a vendor's (Spamhaus). One hit is enough to get listed.

Recycled traps used to belong to real people. The address was abandoned, bounced for months, then reclaimed by the provider as a trap; the mailbox flips from "5xx user unknown" back to "250 OK" and you score a hit on what now looks like a perfectly valid address. This is the most common flavor and the one hygiene prevents directly. Suppress hard bounces on the first failure and you close the door.

Typo traps sit on lookalike domains (gmial.com, yahooo.com, hotnail.com) waiting for senders who never validated a signup form. Real-time validation at the point of capture, meaning domain typo correction plus a DNS MX lookup, stops nearly all of them before they get stored.

How often to clean

Most lists want a full hygiene pass every three months plus a continuous cleanup that runs automatically on every send. Daily and weekly senders lean toward monthly; a monthly newsletter can stretch to quarterly without much penalty. Annual cleanups are theater. By the time you run one, your bounce rate has been dragging reputation down for months.

The continuous piece is what teams miss. Hard-bounce suppression belongs on the first bounce, role-based filters belong at signup, and typo correction belongs in the form rather than a quarterly review. The deep-clean then does the work automation can't: long-term inactives, re-verification of dormant segments, and a look for patterns nobody caught.

A rough heuristic I give customers: if your 30-day bounce rate is above 1%, you're overdue for a full clean; above 2% and you're already in trouble. Track list growth rate next to it, because a list that grows and decays at the same speed looks stable on a dashboard while quietly getting worse.

Verification tools and what they can't do

Verification tools check each address against syntax rules, DNS and MX records, an SMTP handshake, role-based patterns, and known trap databases. You upload a list, the tool returns valid, invalid, risky, or unknown per address, and you suppress or segment on the result. They're useful. They're also imperfect, and the failure modes are predictable enough to plan around.

Verification doesn't fix permission; a bought list that passes verification is still a bought list and will still generate complaints. Catch-all domains, common in B2B, come back "unknown" at high rates, so auto-suppressing them deletes valid corporate contacts. Every check is a point in time, which is why continuous hygiene beats a one-off scrub. And some tools flag engaged recipients as "risky" on domain reputation quirks alone, so cross-check against your own engagement data before you delete anyone.

Run verification on any list you haven't mailed in 60 days or more, before importing from an acquisition or migration, and on any signup source whose quality you doubt. Start with whatever your ESP already does on every send; add a standalone verifier when you need a deep clean on a dormant file rather than as a second permanent subscription. What verification can't do is substitute for collecting permission in the first place.

Handling inactive subscribers

Inactive subscribers get one chance to come back, then they get suppressed. Define the inactivity window (90, 180, or 365 days depending on cadence), pull the segment, run a short re-engagement series, and remove everyone still silent at the end. That's a sunset policy, and every deliverability team I've worked with runs some version of it.

Two orders work: re-engage then sunset, or sunset then re-engage a tighter segment. I prefer re-engaging first, because it's fairer to the subscriber and it recovers revenue you'd otherwise write off. Most of a dormant segment won't come back, and that's fine; those people were already gone, and you're only making it official. The win-back subject lines that work tend to ask a direct question rather than perform sadness about the relationship.

The full sequence lives in the re-engagement guide. Pair it with proper segmentation so the offer matches what the subscriber signed up for; nobody re-engages with a generic "we miss you."

If a subscriber won't open three consecutive re-engagement emails spaced a week apart, they're done. Suppress them and keep the suppression record, so you don't re-import the same address through a form six months later and start the cycle over.

Mistakes that tank deliverability

Five of these turn up constantly in account audits, and each one is avoidable.

• Mailing hard bounces more than once, which happens whenever a platform retries without auto-suppressing. Audit the setting today; it takes a minute.
• Buying or renting lists. Even a verified purchased list violates consent, complains at multiples of an opt-in list, and almost always contains traps.
• Treating role-based addresses like personal inboxes. info@ is never going to read your newsletter the way a named subscriber does; segment them out of bulk sends.
• Skipping signup validation. A form with no syntax check and no double opt-in is a pipeline for typos, traps, and bots. Double opt-in costs you a slice of raw signups and pays it back in engagement quality within a quarter.
• Removing bounces and ignoring engagement. A dead-but-valid mailbox, the one belonging to somebody who quit their job without closing the account, never bounces. It just silently drags your averages down.

One more worth naming: running a giant one-off clean after years of neglect and expecting reputation to snap back. It won't. Reputation systems at Gmail, Outlook, and Yahoo average over rolling windows of 30 to 60 days, so a cleanup removes the cause while the signal takes weeks to fade. Keep sending consistent volume to your engaged subscribers during that window, because a sudden volume drop confuses the same systems you're trying to reassure.

Whether you write it down in a document or a spreadsheet doesn't matter much. Having a policy that someone actually follows is the part that separates senders who stay in the inbox from senders who wonder where their open rates went.

For the wider deliverability picture, the email deliverability guide is the pillar for this cluster; the bounce rate deep-dive covers the numbers, and how to avoid the spam folder ties hygiene back to the placement outcome you're actually after.