Many third-party Exchange backup products include the ability to back up and restore individual mailboxes (also known as brick-level backups). While single-mailbox restoration is appealing, most experienced Exchange administrators don't use brick-level backups very often.
The primary reason for this lack of popularity is that brick-level backups usually take much longer to complete and consume more space on backup media than standard backups. These issues make brick-level backups undesirable or even impossible for most daily backup needs. But when you're dealing with a corrupted database, a brick-level backup can be a real help.
A normal backup of an Exchange database is basically an all-or-nothing proposition. If the backup program detects database corruption, the backup aborts—an intentional decision by Microsoft. Administrators who perform a backup using the standard API can be reasonably sure that they actually have a good backup. In most situations, this is a good thing. Error checking is another reason why administrators generally prefer standard backups to brick-level backups.
Brick-level backups don't check for corruption the way a standard backup does, and they continue to back up mailboxes and messages even if they detect corruption. While this is less desirable in normal circumstances, you can use this less stringent method to your advantage when recovering a corrupted database. You can often salvage information that would otherwise be lost. This can come in handy, especially if you didn't adequately back up the databases before corruption occurred.
While you can use the ExMerge utility much as you would use a brick-level backup, it isn't as convenient as using a good third-party brick-level backup.