Email List Decay: How Fast Your List Rots and What to Do About It

An email list is not an asset that sits still. It is a depreciating one, and the depreciation is continuous rather than occasional. The address that worked last quarter belongs to someone who has since changed jobs, and nobody sent you a notification.

The commonly cited figure is that lists decay at around 2% to 2.5% per month, or somewhere near 25% to 30% a year. That number is useful as an order of magnitude and misleading as a plan, because the variance between list types is much larger than the average suggests.

01What decay actually consists of

Four separate processes, with different rates and different fixes.

People change jobs. This is the dominant driver for B2B lists and it is relentless. Ordinary job-market turnover means a meaningful slice of any professional list becomes invalid every year, because [email protected] stops existing when firstname leaves. Nothing you do to your list prevents this. It is happening in the world, not in your database.

Domains lapse or change hands. Small businesses close. Companies get acquired and consolidate onto the parent domain. A domain expires and gets bought by someone who parks it. In each case the MX record may still resolve, which is why a syntax check tells you nothing here.

Mailboxes get abandoned. Consumer addresses in particular. Someone stops using a Yahoo account they have had since 2009. The provider eventually reclaims it. Reclaimed addresses are also where recycled spam traps come from, which is why the abandoned population is disproportionately dangerous relative to its size.

Addresses were never good. Typos at capture, disposable addresses used to get a download, deliberate junk entered to bypass a gate. This is not decay, strictly. It is acquisition debt, and it was in your list from day one. It matters here because it is indistinguishable from decay when you look at a bounce report.

02Why the rate varies so much

The 2% average hides ranges that differ by a factor of five or more.

B2B lists decay faster than consumer lists. Job changes affect every professional address; nobody changes their personal Gmail because they changed employer. If your list is business addresses, the top of the range is your baseline, not the middle.

Lists acquired through lead magnets decay faster than lists from purchases. Someone who typed an address to get a PDF had a weaker incentive to type their real one than someone who entered payment details.

Frequently mailed lists appear to decay more slowly. This is partly an illusion and partly real. If you mail weekly, dead addresses surface as bounces immediately and get suppressed by your ESP, so your list stays cleaner without you doing anything deliberate. If you mail quarterly, three months of decay arrives all at once and looks like a crisis.

Sector matters. Startups and agencies churn staff quickly. Government, education and healthcare addresses are considerably more stable. A list of university staff behaves nothing like a list of Series A founders.

The practical consequence is that you should not budget your cleaning cadence off 2%. You should measure your own rate once and then plan against that.

03How to measure your actual decay rate

This takes two verification passes and some patience, and it is worth doing once.

Take a fixed sample of 500 addresses from your list. Not a fresh segment, not your most engaged people. A random sample, and write down which addresses are in it, because you need the same ones later.

Run them. Record how many come back deliverable, risky, undeliverable and unknown.

Wait ninety days. Run the exact same 500 addresses again.

The difference in the deliverable count, divided by three, is your monthly decay rate on that list. Not the industry's. Yours.

The free plan is 500 credits a month, which is exactly one pass of this sample, so the measurement costs nothing beyond two months of a free account and remembering to come back.

Most people who do this find one of two things. Either their rate is well under 1% a month, in which case they can stop worrying and clean twice a year. Or it is over 3%, in which case they have a capture problem rather than a decay problem, because genuine decay rarely runs that hot.

04Planning cleaning around the rate you measure

Once you know your number, the cadence follows from your bounce tolerance rather than from a rule of thumb.

Work backwards. Decide the bounce rate you are willing to hit on a send. A commonly used ceiling is 2%, and under the current bulk sender requirements you want to be well inside that. Divide your tolerance by your monthly decay rate, and that is roughly how many months you can go between cleans.

At 2% decay and a 2% tolerance, you clean about every month. At 0.8% decay and the same tolerance, every two to three months is fine. At 4% decay you are cleaning before every significant send, and you should also go and look at your signup form, because something upstream is broken.

This is a more useful frame than "clean quarterly," which is advice calibrated to nobody's list in particular. Our post on how often to clean an email list works through the cadence question by sender type in more detail.

05The thing decay does not explain

There is a failure mode where someone measures decay, sets a cleaning cadence, executes it faithfully, and their engagement metrics keep sliding anyway. The cleaning is working. The list is full of live, deliverable addresses. And nothing improves.

That is because decay and disengagement are unrelated problems that look similar in a dashboard.

Decay is addresses becoming invalid. Verification finds it and cleaning fixes it.

Disengagement is real people at real addresses who stopped caring. Every one of them verifies as deliverable, correctly, because their mailbox exists and accepts mail. Verification has no opinion about whether they want to hear from you, and cannot have one. That signal lives in your ESP's open and click data, not in an SMTP response.

If your deliverable rate is high and your open rate is falling, you do not have a decay problem. You have a relevance problem, and no amount of list cleaning will touch it.

06What to do with the results

When you run a full clean, you get four buckets and only one of them has an obvious action.

Undeliverable comes off the list. This is unambiguous.

Deliverable stays. Also unambiguous.

Risky and unknown are where judgment lives. These are addresses where the evidence is genuinely inconclusive: a catch-all domain that accepts everything so acceptance proves nothing, a server that would not complete the conversation, a role-based address that exists but belongs to a shared inbox. Sending to them is a choice with a cost, and the right call depends on whether you are protecting a fragile sender reputation or running an outbound campaign where a slightly higher bounce rate is acceptable.

We keep those two statuses separate rather than collapsing them into a score precisely because the correct action differs. Our post on what each verification check actually does covers what produces each verdict.

07The short version

Assume your list is losing addresses every month whether you look or not. The average is around 2%, your actual rate is probably not the average, and measuring it takes two passes ninety days apart. Set your cleaning cadence from your own number and your own bounce tolerance rather than from a rule of thumb. And if cleaning is not improving your engagement, the problem was never decay.

Run your first sample and start the ninety-day clock.