Your Email List Decays 23% a Year. Most Teams Find Out Too Late.

Email lists decay at roughly 23% per year, about 2.1% of contacts going bad every month, according to ZeroBounce's 2026 decay report and the long-cited HubSpot database decay model. That means a list of 50,000 you built last January is closer to 38,500 deliverable today, even if nobody unsubscribed. The number to watch is not your subscriber count. It is how many of those addresses still reach a real inbox.

Most teams never measure this. They watch the list grow, feel good about the line going up, and quietly send to a pile of dead addresses, spam traps, and people who changed jobs eighteen months ago. The growth chart hides the rot underneath it. And the rot is what gets your sending reputation downgraded.

The 23% you lose every year is not evenly spread

The headline decay number moves around. ZeroBounce logged 23% in 2025, down from a 28% peak in 2024, with 25% in 2023 and 22% in 2022. So somewhere in the low-to-high twenties is the baseline you should assume, not zero. MediaPost ran a piece literally titled your list decays by 28% per year, and that figure is not an outlier. It is the upper end of normal.

But the average lies to you, because decay is not uniform. B2B lists rot faster than consumer lists, mostly because of job changes. When someone leaves a company, their work address dies, and you usually have no idea it happened. ZeroBounce found that only 62% of all the emails submitted to them in 2024 came back valid. Read that again. Nearly four in ten addresses that companies thought were good were not.

If you run a SaaS or B2B list, assume you are at the worse end of that range, closer to 28-30%, and plan for it. If you are mostly B2C with personal Gmail and Outlook addresses, you decay slower because people keep personal addresses for years. The mistake is applying one blanket assumption to both. From what I have seen, teams that segment B2B and B2C and track decay separately catch problems a quarter or two earlier than teams watching a single blended number.

How to actually measure your own decay rate

You do not need a tool to get a first read on this. You need two numbers and a subtraction.

Pull your hard bounce rate over the last 90 days and annualize it. If 0.5% of your sends hard-bounced last month, that is roughly 6% a year of confirmed-dead addresses, and hard bounces are only the addresses that announce they are dead. The real decay rate is higher, because a lot of abandoned addresses accept mail forever without anyone reading it. Those do not bounce. They just sit there dragging your engagement rate down.

So the better proxy is engagement decay. Take everyone who has not opened or clicked in 90 days, divide by your total active list, and you have your soft-decay percentage. If that number is above 30%, a third of your list is functionally gone whether the address technically works or not. I would treat anything over 25% as a flag worth acting on this month, not next quarter.

One practitioner case documented by Mailtrap is worth holding onto as a benchmark: a sender who pruned inactive contacts saw deliverability climb, open and click rates roughly double, and the email bill drop 47%. That last part is the one nobody talks about. You are paying your ESP to send to people who will never read it, on most platforms by contact count. Dead weight is a line item.

Opens stopped being a decay signal back in 2023

If you are still defining an inactive subscriber as someone who has not opened in X days, your numbers are wrong, and they have been wrong for a while. Apple's Mail Privacy Protection pre-loads tracking pixels for iOS and macOS Mail users, which fires an open event whether a human looked at the message or not. That inflates a big chunk of reported opens, by most estimates around half of them depending on your audience mix.

So a list can look healthy on opens and be rotting underneath. The open rate is partly a robot opening mail on behalf of people who stopped caring months ago. We covered the deeper version of this when Gmail's AI inbox started changing what "in the inbox" even means, and the takeaway carries over here. The signal you trust most is the least reliable one you have.

Clicks are the signal that survived. A click is a deliberate action a pixel cannot fake. Rebuild your engagement definition around clicks first, with opens as a weak secondary input. Yes, click rates are lower and the segments will look smaller and scarier. That smaller number is the true one. Better to act on an honest 15% engaged segment than an inflated 45% that includes a fleet of Apple servers tapping your pixel.

The sunset policy is the only fix that scales

Cleaning a list once is a chore you will not repeat. A sunset policy is the version that runs itself: a standing rule that moves people out of regular sends after a set period of silence, with one re-engagement attempt before they go. Mailgun and Mailflow Authority both lay out the thresholds, and they scale to how often you send.

The rough map: daily senders sunset at 60-90 days of no clicks. Weekly senders at 90-120 days. Monthly senders at 150-180 days. The logic is simple. The more often you hit an unengaged address, the faster it poisons your reputation, so the faster you cut it. A daily newsletter cannot afford to keep mailing someone who went quiet four months ago. A monthly sender has more runway.

Here is the sequence I would actually build, and it is boring on purpose:

• At 90 days of no clicks, move the subscriber into a re-engagement flow. Two or three emails, a clear "still want this?" ask, an easy one-click stay button.
• If they click anything, they go back to the main list. Reset the clock.
• If they ignore the whole re-engagement series for 14 to 30 days, stop sending to them entirely. Do not delete yet. Suppress.
• Run a verification pass on the suppressed pile every quarter to separate truly dead addresses from merely quiet ones.

The reason you suppress rather than delete immediately is that quiet is not the same as gone. Some people read every email in preview without clicking and would notice if you vanished. Suppression lets you keep the record while protecting your reputation. Deletion is for the addresses verification confirms are dead.

What clean actually buys you, in numbers Gmail enforces

This is not hygiene for its own sake. There is a hard line, and it is enforced. Under the Gmail and Yahoo sender requirements, any sender pushing 5,000 or more messages a day to consumer inboxes has to keep their spam complaint rate below 0.3%. Google's own guidance says aim for under 0.1%, because 0.3% is where enforcement starts, not where you are safe. Mailgun's breakdown puts it plainly: send 10,000 emails, collect 30 complaints, and you have hit the ceiling.

A rotting list is a complaint machine. People who forgot they subscribed hit the spam button instead of unsubscribing, and each one of those counts against you with the inbox providers. The decayed portion of your list is exactly the group most likely to mark you as spam, because they no longer remember why they are hearing from you. So decay does not just waste sends. It actively manufactures the complaints that get you filtered.

And if your authentication is shaky on top of a dirty list, you compound the problem. Worth getting the foundation right first: we walked through why SPF, DKIM, and DMARC can all be set up correctly and you can still get rejected. Authentication plus list hygiene are the two halves of staying in the inbox. One without the other does not hold.

The mental model I keep coming back to is a swimming pool. Decay is evaporation. It happens constantly, quietly, whether or not you are paying attention, and if you only ever top the pool up without checking the water, you eventually find yourself swimming in something you would not drink. Acquisition is the hose. Hygiene is the filter. Most teams run the hose at full blast and never turn the filter on.

FAQ

How often should I clean my email list?

Run a real verification pass quarterly, and let an automated sunset policy handle the day-to-day in between. Quarterly catches the accumulated dead addresses that did not bounce, while the sunset rule stops unengaged contacts from piling up in the first place. Cleaning once a year is too slow when a quarter of the list turns over annually.

Does removing subscribers hurt my reach?

It does the opposite, and this is the counterintuitive part most teams resist. A smaller engaged list lands in the inbox more reliably than a big list full of dead weight, because inbox providers read engagement as the signal of whether you are worth delivering. Sending to 10,000 people who open beats sending to 50,000 where 35,000 ignore you. The second list looks better on a slide and performs worse in reality.

What counts as inactive in 2026?

No clicks, not no opens. Because Apple Mail Privacy Protection fires open events without a human involved, opens overstate engagement by roughly half for many lists. Define inactivity by the absence of clicks over your sunset window, treat opens as a weak supporting signal, and your "active" segment will finally reflect people who actually read you.

If you do one thing this week, calculate your soft-decay number: everyone with no clicks in 90 days, divided by your total list. If it clears 25%, you are not running an email program, you are funding a graveyard. Build the sunset rule before you send the next campaign, not after the next deliverability scare.

Notice Me Senpai Editorial