Let’s talk about (one of the) worst naming decisions in tech: what is known as unstructured data.
It’s not “the cloud”. It’s not even “Big Data”
It’s “unstructured data”
This term confuses people, wastes time to describe what it is, and is just generally wrong.
Structure is Everywhere
There’s no such thing as unstructured data.
Every piece of digital information has some kind of structure. Even that random video file your marketing team uploaded has metadata – creation date, file size, codec information, duration.
That’s structure right there.
A Word document? It’s got paragraphs, headers, formatting. Structure.
An email? Subject line, sender, timestamp, body text. More structure.
A photo has EXIF data telling you when it was taken, which camera was used, and most likely GPS coordinates.
The problem isn’t that this data lacks structure – it’s that the structure doesn’t fit neatly into rows and columns like a database.
What “Unstructured” Really Means
When people say “unstructured,” what they actually mean is “doesn’t sit nicely in traditional databases”
It’s data that can’t be shoved into the rigid rows and columns that make database administrators happy. That doesn’t mean it’s unstructured – it just makes it different.
Think of it this way: jazz improvisation isn’t “unstructured music” because it doesn’t follow the same rules as a classical symphony.
Brief aside – as a lifelong jazz lover (and aspiring musician) there’s so much “structure” and musical theory and beautiful math that goes into improvisation it’s insane. But that’s for another blog post
Oh, yes. The music metaphor also applies to data.
Chaotic, Messy or Untapped?
Calling it “unstructured” makes people think it’s chaotic, messy, or somehow less valuable than “proper” structured data.
It’s in fact the opposite; within unstructured data there is every company’s intellectual property, almost limitless value and huge opportunity for enterprises.
Thanks, Evil Computer Overlords.
The AI revolution has helped people realise that there are tools that we have that can finally bring structure to this unstructured world.
The AI Opportunity
AI doesn’t care about database schemas or whether something fits in a spreadsheet.
Large language models can read through thousands of unstructured documents it has never seen before and pull out insights that would take humans weeks to find.
Computer vision can analyse video content and describe exactly what’s happening in each frame, recognise objects and even the mood of each scene.
Machine learning can spot patterns in support tickets that reveal how customers are feeling about products.
Suddenly, all this “messy” unstructured data becomes valuable.
Companies that figure this out first aren’t just managing their unstructured data better – they’re turning it into a competitive advantage.
What Actually Matters
This blog was born from a tirade against one of the many unsuitable naming conventions in technology.
But the message here is that unstructured data powers the biggest businesses worldwide. It helps companies analyse huge datasets to improve healthcare outcomes, make energy generation more efficient and generally make the world a better place. (Just in time for the robots to take over)
Unstructured data is awesome. What’s not awesome is the fact that it’s often not properly managed or controlled.
The Real Structure
Good data management isn’t about forcing everything into rigid rows and columns. It’s about understanding the structure that exists and working with it.
If it needs to be in workflow-centric folder structures that’s cool.
If it needs to be under the management of a niche application, that’s also cool
Managing all data is important, so when a business has a gold mine of “unstructured data”, it needs the right tools to extract that gold.