Why You Need to Back Up Your Research Locally

Think about how much time and effort you’ve invested in your research.1 And think what you would have to do to recreate it all if your research instantaneously disappeared.

If you’re like me, thinking about mass deletion of your research probably has your stomach turning.

Keeping your work in “the cloud” has the advantage of ensuring it doesn’t just live on one computer of yours. But at the same time, nobody has more incentive to keep your research safe than you do.

That’s definitely true for free cloud storage. But it’s still true even if you’re using paid cloud storage.

Paranoia Can Be Productive

To think about a major cloud storage provider failing and losing your research might border on paranoia. But in your work, there’s a kind of paranoia that’s paralyzing. And there’s another kind that’s healthy and productive.2

That healthy and productive kind of attention leads you to see the uncertainty that exists. It helps you avoid succumbing to the hype that nothing can go wrong.

It’s the kind of attention that, having seen this uncertainty, doesn’t lead you to panic or perpetual dyspepsia. Instead, it’s the kind of concern that leads you to soberly identify and adopt ways of limiting your risk. By doing so, you create an insurance policy against the unexpected.

Can Your Research Really Disappear?

But is this an actual danger?3 Is it really feasible that you might lose your research? In a word, yes.

If you have printed material, that can get consumed in a fire, left on public transit, (despite the cliché) eaten by a dog, or meet any number of other sad ends.

But even if you have things “safe” in the cloud, changes might fail to upload properly, or the most current version of a file might get deleted or copied over.

Take the example that John Meade has shared about the nightmare scenario visiting his OneDrive:

I guess the new version of my #sblaar18 #aarsbl18 paper is better than my old one. Thanks @Microsoft for deleting previous versions of my paper and not allowing me or IT to recover them in order to improve my work. 😳🤨

…. Unfortunately, my institution uses OneDrive which has taken years away from me over the past 48 hours. I have very important stuff saved to Dropbox but I don’t have a lot of space left on there :-(.4

Similarly, I’ve had to dig way back into backup sets to recover material on a couple of occasions. And “rewind” features in cloud services haven’t always proven reliable. In fact, in one case, I’m reasonably sure using this “feature” simply made the recovery process more convoluted than it otherwise would have been.

That said, I’ve thus far always finally found the missing pieces, but only because I had local backups that went back farther than the versions saved in my cloud services. Finding those pieces took a while, but not nearly as long as it would have to recreate the material that went missing.


So, in the end, yes, losing research you’ve labored over is a real danger. Cloud services can be part of the solution and do provide convenient ways to back up your work in multiple locations.

But they aren’t sufficient insurance against this danger. Instead, you also need to have a local backup as well, and when you do, your research will be that much safer.

Depending on how you assess the risks, you can leave this drive connected in order to make constant backups. Or you can unplug it between periodic backups so that its contents can’t get overwritten unintentionally. Macrium Reflect, the software I use for backups, even has an “Image Guardian” that will prevent illegitimate overwrites of your backups from drives that are connected.

In the end, there’s no completely failsafe way of protecting yourself from losing a painful chunk of your research. But nowadays, it’s incredibly easy to drastically reduce the likelihood that you’ll experience a significant problem and also turn any difficulties you do have from occasions for lament into comparatively minor inconveniences.

  1. Header image provided by Markus Spiske. ↩︎
  2. For further discussion of the concept of productive paranoia, see Jim Collins and Morten T. Hansen, Great by Choice: Uncertainty, Chaos, and Luck—Why Some Thrive despite Them All (New York: HarperCollins, 2011). ↩︎
  3. Image provided by Elisa Ventur. ↩︎
  4. Tweeted by John Meade on 1 November 2018; emphasis added. ↩︎

Some of the links above may be “affiliate links.” If you make a purchase or sign up for a service through one of these links, I may receive a small commission from the seller. This process involves no additional cost to you and helps defray the costs of making content like this available. For more information, please see these affiliate disclosures.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.