Strength in numbers: essential tools for analysing your business data
Strength in numbers: essential tools for analysing your business data
By Stephanie Oley
Big data can deliver powerful advantages to business – such as sharper customer targeting, better market insights, and improved operational efficiency. But it’s also unwieldy.
When your datasets contain unthinkable amounts of data, then it’s critical to have the right skills and tools to interpret this information. In this article, we’ll introduce some of them, along with examples of the insights they’re delivering to organisations.
Your business data: Structured or unstructured?
Let’s start by considering the data you have at hand, and how structured (or usable) it is. Structured data is information already organised into predefined fields, such as in a spreadsheet. For example, a sales report will have fields showing sales by region, product, date, salesperson and so on. Health records will have fields for names, contact information, medical history and dozens more factors. This was once the most common form of data.
However, today’s data can be semi-structured, with emails being one example. Emails have fields for a limited amount of information, such as the sender and recipient’s names, addresses, and date sent. However, much information remains unknown – such as the email’s purpose, antecedent, level of urgency and other factors.
The Internet of Things has also led to an explosion in the volume of unstructured data. This is raw digital information that often does not even fit into a spreadsheet. Social media photos, website videos, data from intelligent devices, and conversation transcripts are all examples of unstructured data.
Not surprisingly, this category of data is estimated to be growing much faster than the other two.
Harness the power of unstructured data, though, and businesses can make a raft of powerful decisions. For example, a marketer could analyse images of restaurant meals posted on social media, to better understand what’s hot and what’s not. A public health official could harness health records to understand the resources needed by specific communities – whether it’s combatting domestic violence, or treating coronary health.
Tools for analysing big data
How can your business extract meaning effectively from such vast and varied sources of data?
Start by ascertaining your team’s skillset in the two main skillsets involved. One is data science, which involves cleaning and massaging the unstructured data to make it usable. The other is data analytics, which involves drawing insights from that data.
Below are four readily available tools, two of them open-source and free. They are just some of the tools and techniques explored in CCE’s Business Analytics for Large Data Sets course.
- Orange Data Mining – An open-source data science tool that can visualise data in various formats, from standard statistical distributions and scatter plots, to sophisticated decision trees and heatmaps. Orange can work with much unstructured data, too. For example, it can cluster similar images together to help with scenarios such as the marketing challenge described earlier.
- Exploratory.io – An open-source data science tool that is actually the front-end tool for the R programming language. Exploratory allows users to extract insights from various types of data using a simple point-and-click method.
- PivotTables and PivotChart – A Microsoft analytics tool that’s like an infinitely more powerful version of Excel, able to filter, group and reorganise large datasets to meet specific objectives. Users can ‘pivot’ different rows and columns to see different information groupings, presented as a table. You can then use the charts function to visualise those results.
- Tableau and Power BI – Two popular paid tools that can clean and integrate data from multiple sources. Analysts who are already comfortable within the Microsoft suite may prefer the compatibility of Power BI. By contrast, those seeking a more powerful option may prefer Tableau by Salesforce, which can crunch much larger volumes of data.
Which organisations are using big data to effect?
Data is now being generated by an increasingly wide range of sources, such as intelligent devices, social media, and better-networked computers. This is helping organisations in Australia and overseas to make large strides in service delivery.
- Transport networks – Every time users tap on and off their bus or train using a contactless smartcard, they’re sharing data that benefits others. For example, public transport operators now draw on this information to plan timetables that meet genuine demand patterns
- Car manufacturers – The Internet of Things (IoT) is allowing car manufacturers to analyse the data transmitted by their vehicles, to better allocate resources. For example, vehicle data can reveal which spare parts are required in different geographies, enabling just-in-time deliveries.
- Public health organisations – For example, clinical trials can analyse respondents’ behaviours to identify who is most likely to persist with a study. Another discipline that uses data effectively is gene therapy, which is already helping to determine susceptibility to different diseases.
The volume of data generated worldwide is growing exponentially, and the data your business will generate by 2023 could well be 10 times the volume of today.
Whether you’re in marketing, technology, human resources, logistics or another sector, gaining the confidence to analyse large data sets will help your business deliver transformative results.