This article was originally published in Dutch in 2012 and has been translated and updated in October 2026.
A term heard more and more often among IT professionals and (large) companies is ‘Big Data’: gigantic collections of data that cry out for structure, structure that traditional analysis software can no longer bring to the mess. Companies that create order out of this chaos can benefit from it and stay one step ahead of the competition.
Big Data?
So what exactly is Big Data? Big Data is not just ‘big’ because of its sheer volume, but because of three related factors: volume, velocity and variety. Analysing it properly requires more advanced software than the programs that were common, or ‘traditional’, at the time of writing. Without it, making connections is simply impossible.
Structured and unstructured
Big Data can come in structured form (think of retail transactions per shop or energy readings per household) as well as unstructured form (think of blogs, status updates on social platforms, photos, and reviews of products and services). Whereas structured data is relatively easy for an organisation to store in databases, unstructured data is much harder to capture. When Big Data is analysed well, it gives organisations insight into their operations. They can then use the resulting information to make their processes run more smoothly.
Big Data explained on video
Big Data and cloud computing specialist EMC was increasingly focusing on analysing Big Data for businesses at the time. In a video, EMC CTO Patricia Florissi explained in unmistakably Italian-accented English what Big Data actually is. One of her conclusions was that it is not only the size of the data that counts, but also the speed at which it is generated and the many different sources it comes from. As an example, she compared a 40 MB PowerPoint presentation with a one-terabyte (TB) X-ray image. The terabyte is obviously larger, but if the PowerPoint cannot be sent by e-mail and gets in the way of employees doing their work, those 40 MB suddenly become very big indeed. Big Data, according to Florissi, covers anything that tests and exceeds the limits of existing systems.
What exactly can companies do with Big Data?
Companies and individuals love storing data. According to EMC, the amount of information stored worldwide in 2011 was heading towards 2 zettabytes (2 trillion gigabytes), nine times as much as five years earlier.
You can imagine that analysing large streams of data is becoming ever more important. Just look at the explosive growth in data traffic on social networking sites such as Twitter and Facebook. Users upload millions of photos and send billions of updates about their lives and the latest news. At the end of 2011, every minute saw 98,000 tweets, 695,000 Facebook updates and 60 hours of video uploaded to YouTube. That is an invaluable source of information for companies that want to reach their customers in a more targeted way. And it is not just social media that deals in Big Data: financial transactions and weather data also generate an enormous stream of data that can yield valuable information. If that information is then used well to improve business processes, companies can potentially turn it into money.
Big Data in practice
Big Data was already being used successfully by several companies, such as the retail chain Wal-Mart and the telecoms provider UPC Cablecom.
Example: Wal-Mart
The absolute pioneer of commercial Big Data applications is the American retailer Wal-Mart. Wal-Mart’s inventory management system tells suppliers exactly how sales of their products are going. Several large suppliers have even been made fully responsible for stock management and only sell their product at the moment it passes through a Wal-Mart checkout. In this way, the supermarket has been turned into a highly efficiently run warehouse.
Example: UPC Cablecom
Swiss telecoms provider Cablecom noticed that many customers cancelled their subscription after thirteen months. Analysis showed that the customer service department was already receiving significantly more calls from these customers in their ninth month. Cablecom responded by offering customers a special deal for renewing their subscription in the seventh month. According to the figures reported at the time, this reduced the share of customers leaving from 20 to 5 per cent within a single year.
Collaboration between IT and other departments
The second example in particular shows effective collaboration between different departments: IT (data analysis), Customer Service (registration) and Marketing (initiating and/or responding to analyses, and making the right offer at the right place and time).
In the Netherlands, several companies were already experimenting with Big Data, such as Bol.com and FBTO, part of Centraal Beheer Achmea. The latter had been experimenting with Big Data for quite some time. A few starter tips from FBTO’s analysts:
- Start small: begin by experimenting with data analysis;
- Take the human factor into account: take your time. Involve people from different departments. Build up knowledge. Find out what your organisation’s questions and objectives are.
Tips for getting started with Big Data
Companies that want to do more with the gigantic mass of unstructured data would do well to develop a Big Data strategy. A Big Data strategy allows you to make market predictions based on rapid analysis, and in that way gain a competitive advantage: more profit, lower costs. When you start, keep the following in mind:
1. Take your business objectives as the starting point. Determine which data you need to draw the right conclusions about your business objectives. As a marketer, you can take the derived marketing and campaign objectives as your starting point and define what information you need to determine whether your marketing activities are delivering the desired results;
2. Take stock of what marketing information is already available within the organisation and what information is still missing.
3. Establish the relationships between different data. Give meaning to certain combinations and act on them. Define ‘if-then’ scenarios. That way you can switch quickly and take action when the system detects a particular ‘if-then’ outcome. Establishing these relationships between data therefore has to be done extremely carefully, as everything depends on it;
4. Good collaboration between IT and Marketing is essential. Both Marketing and IT can drive the successful use of Big Data. IT analyses, but Marketing has to do something with the results. Conversely, at Marketing’s initiative, IT can provide insight into how campaigns are performing, so that results and customer satisfaction can be brought to and kept at the desired level. It is not easy, but as a marketer you can also try to gain insight into the buying process by investigating and testing the relationship between campaigns, time-to-conversion and touchpoints (websites, social media).
5. Technical infrastructure and knowledge. Big Data will not be a success without the right knowledge of systems, software and data among IT staff. They need to be able to map out which data exists and technically bring together dissimilar, structured, unstructured and sometimes even stand-alone systems (think of payroll, customer management, marketing campaign and ‘social listening’ systems). The IT infrastructure needs to be reviewed and the right software deployed. In the end, the systems must produce the right reports, on the basis of which correct and fast decisions can be made and/or set in motion automatically.
Will Big Data really become big?
Yes. Millions of dollars and euros were already being poured into software for analysing Big Data, and a Google search for jobs related to Big Data was already turning up job titles such as ‘IT architect Big Data’. These Big Data scientists will increasingly be needed to carry out the complex analyses involved. In addition, targeted analysis of customer data can give companies a more focused approach and strategy, allowing them to serve their customers better. This also applies to the Marketing department. Combining it with data from social media can enrich the knowledge base enormously and reveal interesting insights.
A word of caution
The line between what is private and what companies are allowed to run off with is a thin one, however. Do all those millions of Facebook users give their data to the platform, or to companies that are allowed to use it for marketing purposes? Are companies allowed to simply phone consumers who are browsing their site with sales pitches? Interesting questions that were bound to be debated at length in the years to come.