The data value chain
1. The data value chain
When you think about data processing, it's important to place it within the broader context of the data value chain. Imagine data traveling along an assembly line, like a car in a factory. The assembly line progressively adds parts and value to an object that moves along it. Raw data at the beginning of the line is eventually transformed into actions that humans or machines take. Let’s examine the steps in this data value chain. Data genesis is the initial creation of a unit of data; this could be a click on a website, the swipe of a card, a sensor recording from an IoT device, or countless other examples. It’s the raw material that will eventually be turned into an insight ready for action. Data collection brings that initial unit of data to the assembly line through ingestion. The basic function of ingestion is to extract data from the system in which it’s hosted and bring it to a new system. It can have dramatically different requirements based on the volume, velocity, and variety of the raw data that’s required for a given analysis, and how fast the data needs to be analyzed. Data processing is where the collected raw data is transformed into a form that’s ready to derive insights from. The data will likely need to be adjusted, for example, by merging different datasets together. It can be a single-stage operation, or it can be a complex tree of cascading procedures. In our manufacturing process analogy, this phase is where raw materials take the shape of the pre-assembly parts of a manufactured product. Data storage is where the data lands, can be found, and is ready for analysis and action. As with real-world manufacturing, where storage options vary depending on the type of product that is processed, different types of data can be stored in different ways. For example, NoSQL is available for fast reads and writes, data warehousing for fast access to analysis, and object storage for unstructured data. There are also customized options of these standard stores. Data analysis provides direction for business-oriented action. To continue with our manufacturing line analogy, in this stage, inputs from the data processing stage are assembled into a final product. And finally, the last step in the data value chain is data activation. When an analysis is produced, it needs to be pushed to the relevant business procedures and decision makers so that action can be taken and the value chain completed. The most common points of activation are applications that make automated decisions and business intelligence dashboards that guide humans toward better, more informed decisions. In our manufacturing line example, this is the step where a fully produced product is put to its intended use. There is no one way to assemble a data value chain, as there’s no one way to create a real-world manufacturing line. Similarly, as technologies progress, new inputs become available, your workforce evolves, or the desired output changes, the optimal value chain will also change. However, at its core, the value chain principles hold. We want to use raw data to perform actions that benefit the business.2. Let's practice!
Create Your Free Account
or
By continuing, you accept our Terms of Use, our Privacy Policy and that your data is stored in the USA.