The strategy I use for keeping data safe follows the following set of guidelines.

A good back up has data that is:

Stored separately from the original copy: If a pipe above your computer were to break, it would destroy both your computer and the backup hard drive you had sitting next to it. So we want at least one of our backups (as the 3-2-1 rule suggests) in a different location. Storing information on the cloud is one approach. Years ago, before computers were a thing, one of the scholars who trained me would send a draft of his book to his friend in Michigan in case his house burnt down. He’d never heard of any natural disaster striking both California (where he lived) and Michigan at the same time so figured it would be a good place to have a backup copy of his work stored.
Backed up automatically: Humans are forgetful creatures. The busier and more stressed we get, the more likely we are to forget to back up. So you want at least one of your backups to work without any input from you. The simplest way to do this is with a cloud back up (one drive, dropbox etc), where your computer automatically backs up all of your pertinent files on a regular interval
Back up regularly: Your backups should happen often enough that you won’t lose anything critical if your computer breaks down. For me personally, my files are backed up either twice daily (datastore) or every single time I save a file (dropbox). Often that means my files are backed up 6-10 time per day. This can come in really handy when I make a mistake. Worst case scenario, I end up losing 20-30 minutes of work. Annoying but manageable.
Versioning: It is important that the online backup program you use includes an option called “versioning.” Versioning means that your backup system keeps several copies of each file. Online programs like dropbox, for example, make a new copy of each of your files every time you click save. In practice, this could mean a single file on your computer might have 30-40 backups, each one reflecting a handful of small changes you’ve made to it. Versioning is important because files can become corrupt or you may dislike a change you made to one of your files and want to be able to go back to an earlier copy of that file. If you save over your old backup with a corrupted version of the new file, your backups are now useless. With versioning, it is possible to go to back to a time before your file became corrupted in the first place.
Utilizes more than one system: Your data protection approach should work even if multiple things to wrong. What if your computer dies and your back up program failed? That would be a disaster. So we want some redundancies. My goal is to have a system where even if my primary storage (my computer) and my backup storage (cloud storage) fail at the same time I lose no more than a few weeks of work. That would still be awful, but I can recover from that. But I’m not sure if I could recover from losing 10 years’ worth of work. So my approach has been to have files stored:
1. 1. On my computer
  2. Off-site via an online program (eg. dropbox, one drive)
  3. Through a university service (datastore)
  4. On multiple hard drives located in different physical locations (family member’s houses in other provinces) that are updated every few months
  5. Various other online services (emailed to myself). I use this for files that are small (word documents) but time consuming to make (a chapter of a book I’m writing)
Test your backups regularly: It is important to verify that the backups you’ve created actually work. You should test your backups every few days/weeks when you first set them up and no less than every few months after that.