Structured vs Unstructured Data: Complete Guide

With all the talk about big data and the ways companies use it, you may find yourself thinking: “What types of data are companies using?” First of all, it is important to understand that not all data is created equal. For example, data from social media channels (such as Twitter, Facebook, etc.) is very different from data gathered from customer relationship management (CRM) or supply chain systems.

Before we start, when you have a basic understanding of the differences between qualitative vs quantitative data, you can then try to make sense of data structures or lack thereof.

So, in short, quantitative data can be counted, measured, and expressed using numbers, for example, height, weight or the cost of something. On the other hand, qualitative data is descriptive and categorial findings collected through, for example, questionnaires, interviews, or observation. Examples of qualitative data could be religion, gender, type of instruction, etc. 

Moving on, in this article we will look at what is called structured and unstructured data. Before we begin, it could be interesting to know that according to Gartner, the majority of the data is unstructured.

In addition to being used, collected, and scaled in different ways, structured and unstructured data will often be stored and managed in completely separate databases. I like this quote from G2 as I think it summaries the differences very well.

I like to think of structured data as data I can store and display in fixed rows, columns and relational databases that are easy to search, compared to unstructured data that cannot be stored and searched as quickly.

Examples of structured data could be dates, customer number, phone numbers, etc. and unstructured data examples could be email messages, customer chats, video files, images, and so on. 

This image summarise the difference between structured and unstructured data nicely.

Image source: Medium

Now that we have got an initial understanding, let’s look a bit closer on structured and unstructured data with some examples.

Structured data is usually classified as quantitative data, and it’s the type of data that we are used to working with. I like to think of structured data as data that fits nicely within fixed fields and columns, for example, in a spreadsheet.

Therefore, structured data is highly organised and can be displayed in rows, columns and relational databases (such as SQL)

Common examples of structured data are Excel files or SQL databases, as each of these has structured rows and columns that can be sorted.

Example of structured data include:

• Age
• Dates
• Currency
• Phone numbers 
• Customer ID:s
• Credit card number 
• Product numbers and ID:s
• Transaction numbers
• Competitor analysis

Unstructured data is information that either does not have a predetermined data model or is not set up and managed in a predefined manner. Unstructured data is typically text-heavy but may include data such as dates, numbers, and facts.

Finding insights within unstructured data can be quite tricky (especially with a large data set), but when appropriately and thoroughly analyzed, the unstructured data can provide some really valuable and interesting insights to extract, for example, customer opinions, employee feedback, expert interviews, etc.

Examples of unstructured data

• Text files such as reports, emails, social media posts, etc 
• Interviews
• Audio
• Video 
• Images

There is also what could be described as semi-structured data, which includes primarily unstructured data, but is vaguely categorized with “meta tags”. Meta tags are snippets of text that describe the content. 

Semi-structured data can easily be broken down into predefined categories, but the information within these categories is, itself, unstructured. In other words, the data set is structured in categories, but the data within each of these categories are unstructured.

An example of semi-structured data could be email, which you can categorise by Inbox, Outbox, Drafts, etc. Another example of semi-structured data could be server logs.

Below are some examples, in no particular order, of software that can be used for structured and unstructured data gathering and analytics.

• SQL for structured data 
• NoSQL (Not Only SQL) for unstructured data
• Microsoft Excel
• Microsoft Power BI
• Salesforce 
• Tableau 
• Apache Hadoop
• Oracle BI
• RapidMiner
• SAS Viya and TextMiner