Written by: 
This is an excerpt of my guest blog entry on osSupportDesk's blog. There's a link to the full article below.
Web site backup comes with its own set of limitations and pitfalls. If you trust your web host for backup you might find your expectations fall short. Most hosts take daily backups – if any at all –on a secondary hard disk on the same server or, even worse, on a secondary partition of the same hard disk. If the server goes down due to a hardware fault, so does your backup. A few enlightened hosts also take backups on remote storage, for example NAS arrays. Even they do so on rather sparse intervals, for example twice per week. This means that on a complete catastrophe you will most assuredly lose a fair amount of data.
The solution is simple in concept. Take your own backups and store them on a cloud storage service, like Amazon S3. Taking your own backups means that you get to decide which data and how often has to be backed up, making sure that the crucial, regularly updated information routinely ends up in a backup archive. Using a cloud storage device adds a strong data safety clause to your procedure, while keeping costs low. Cloud storage is designed to be redundant and reliable, boasting a negligible risk of data corruption or data loss. Combined with its incredibly low cost, it is reasonably attractive to businesses of all sizes: from hobbyists and sole proprietorships up to large corporations and government agencies.