Definition
Big data refers to datasets whose volume, velocity or variety exceed what a single conventional database and a spreadsheet can handle, and to the technologies built to store and analyse them across many machines. The classic definition uses three Vs: volume (terabytes to petabytes), velocity (data arriving continuously, in real time) and variety (structured tables mixed with text, logs, images and sensor readings). The term became popular in the early 2010s alongside tools like Hadoop and, later, Spark.
In a company, big data is the clickstream from a website, transaction logs, sensor data from equipment, support conversations and marketing events, all landing together and needing to be joined. It powers recommendation engines, fraud detection, demand forecasting and, increasingly, the training and grounding of AI models. The infrastructure behind it is the data lake, the data warehouse and the pipelines that move data between them.
The misconception is that big data is a goal. Most companies do not have big data; they have medium data with poor organisation, and a well-designed conventional database or a cloud warehouse handles it fine. In 2026 the term itself is used less, replaced by talk of data platforms and AI readiness, but the underlying discipline is more important than ever: models are only as good as the data behind them, and volume without quality produces confident nonsense.
In practice
A logistics company collects location pings from thousands of vehicles every few seconds. Storing and analysing that stream to optimise routes is a big data problem. Their customer list of 3,000 companies is not, and it lives happily in a standard database.
Why it matters
Big data is a means, not a strategy. Before funding a data platform, ask which decisions would change if you had the analysis, and whether your existing data is even clean enough to trust.
Frequently asked questions
- How big does data have to be to count as big data?
- There is no fixed threshold. The practical test is whether a single machine and a conventional database can store and query it in reasonable time. When you need to distribute storage and processing across many machines, or handle continuous real-time streams, you are in big data territory. Most companies are not.
- Is big data the same as AI?
- No, but they are related. Big data is about storing and analysing large volumes of information. AI, and in particular machine learning, learns patterns from data, and large well-organised datasets make models better. You can have big data without AI, and you can use AI on small datasets.