As we use more automatic notifications, we are running into an increasingly frustrating problem: bad emails. It is my understanding now that the amazon mail-server pushpage.org and schoolnotifications.org are using mark emails bad for up to 2 weeks after just one soft or hard bounce., regardless of the "bad" state indicated in WH. While I can understand a hard bounce, we often have users with full mailboxes, etc that create a temporary soft bounce. This is poorly communicated, and even harder to diagnose. The bounce work-list is cumbersome at best, and is only available for pushpage. We need a better solution to manage bounced email addresses, including a way to remove the bad marker and restore email services to our constituents in a more timely manner. We depend on notifications and pushpages to communicate often very time sensitive information to our constituencies, and have the expectations that mail will reach them. In short, we really need: 1. More transparency about the bounce protocols, including unifying the Amazon mailserver "bad" status with WH. It's frustrating trying to diagnose why a parent is not receiving an email when they are marked bad by Amazon, and not by WH. 2. A updated bounce worklist than enables speedy diagnosis (including return headers) and correction, and includes all modes of website communication including notifications. I'd be curious to know how other schools are handling this issue, and in particular, the resolution of bad emails.