Glossary›Unstructured Data Basics›Dark data

Dark data

What is zombie data?

Zombie data is data that stays on storage after the project, application, or person that created it is gone. It still uses capacity and still appears in backups, but it is no longer read and has no clear owner. Old project folders, copies left behind by a finished migration, and files from people who left the organization are common examples. Zombie data is a large part of dark data, and it can be found from metadata such as last access date, owner, and path.

What is shadow IT, and how does it hide unstructured data?

Shadow IT is technology a team uses without the IT department knowing about it, such as a personal cloud drive, a storage device bought on a department budget, or a cloud bucket opened for a single project. The files on it sit outside the inventory, outside retention rules, and often outside backup plans, so they become dark data from the day they are created. Bringing those systems into one index is what makes the files visible again.

Why is dark data hard to find?

Dark data is hard to find because nothing about it stands out. Four reasons keep it out of sight:

• It is spread across systems. File servers, NAS, object storage, and cloud accounts each have their own tools, and none of them shows the whole estate in one view.
• There is too much of it to review. Large organizations hold millions or billions of files, far more than anyone can open and check one by one.
• The context is missing. Folder and file names such as “temp” or “final_v2” do not say what a file is for, and the people who knew have often moved on.
• Nothing flags it. Dark data causes no errors. It sits quietly on storage, gets backed up with everything else, and only shows up on the bill.

The way in is one index of file metadata across every system, filtered by attributes such as owner, age, file type, and location. Files with no active owner, no recent access, and no clear project are where dark data usually sits.

How do you reduce dark data?

By describing it, one step at a time:

1. Index the metadata of every storage system into one catalog.
2. Assign owners from file owners, project folders, and the company directory.
3. Sort what you find into groups such as active, cold, duplicate, and orphaned.
4. Decide per group: keep, move, archive, or delete, with the owner’s approval.

Dark data shrinks as soon as files have an owner and a purpose on record, even before anything is moved.

← Back to Unstructured Data Basics

Last updated: November 13, 2026.