What are reputation decay models (how fast do filters forget)?
Still have a question, spotted an error, or have a better explanation or a source we should cite?
You fix your sending habits, clean your list, fix your authentication, and then... nothing changes. Deliverability is still tanking. The frustrating truth is that filters don't forget instantly. They run on something called reputation decay models, and how fast your past behavior stops counting against you depends on the provider, the signal type, and how bad the offense was.
Here's what the industry broadly understands about decay windows, with the caveat that no major provider publishes exact formulas. This is built from observed behavior, public documentation, and a lot of painful trial and error.
The general decay window is 30 to 90 days. Most filters weight the last 30 days most heavily. Gmail and Yahoo Mail are known to use rolling engagement windows, where recent opens, clicks, and complaints count more than anything older. A complaint spike from three months ago matters far less than a clean 30-day run. But a complaint spike from two weeks ago still very much matters.
A rough breakdown by signal type:
- Complaint rates: The most immediate signal. A single bad week (above 0.1% at Gmail, or above 0.3% at Yahoo) can start moving your sender reputation in the wrong direction fast. Recovery typically requires 4 to 8 weeks of clean complaint rates, assuming you fix the root cause and don't just wait it out.
- Engagement signals (opens, clicks, replies): These decay within roughly 30 to 60 days. A subscriber who opened regularly six months ago but has gone silent since then contributes less and less to your reputation. Their inaction starts to count against you, not for you.
- Spam trap hits: Slower to recover from. These can affect your sending reputation for 60 to 90 days or longer, especially if the trap belongs to a major data provider. Hitting a Spamhaus-listed trap is in a different league from a recycled address, and it can trigger blocklist entries that have their own separate timelines.
- Blocklist listings: These don't decay automatically. Each blocklist has its own removal process. Some auto-delist after 30 days of clean behavior. Others require a manual delisting request. Spamhaus, for example, won't forget until you actively fix what caused the listing and request removal.
The honest answer for Gmail specifically is that Google uses machine learning across thousands of signals, and their exact weighting isn't public. What is public is that their filtering decisions are heavily influenced by what recipients actually do with your mail. If you can generate positive engagement (opens, clicks, folder moves to inbox) during a recovery period, that speeds things up considerably.
What this means practically: reputation is always moving. You can't build a strong history and then coast on it. Two months of low engagement can undo a year of good behavior, not overnight, but gradually. The inverse is also true. You can absolutely recover from a bad patch if you send better mail to a cleaner list consistently over several weeks.
If you're trying to diagnose a current reputation problem, our free blocklist checker is a good starting point to rule out any active listings. If you're in a genuine deliverability crisis and decay is only part of the picture, the SOS hotline is free and we'll help you work through it.
The most useful email questions, once a month
One email a month. New questions worth reading, real deliverability changes (Gmail and Outlook updates, etc.), and any tool or report that landed since the last issue. No fluff, no upsell.
Contributors
Who worked on this answer
Every name links to their profile. Every company links to their site. Real people, real accountability.