Let's say this straight. We’ve met with dozens of companies that don’t manage their data lifecycle either at all or throughout their lifecycle. This means that when new systems or technical solutions are made, the data is not told how long the data is valid, whether there are statutory retention periods or other factors that affect how long the data can be utilized or how long it is valid for use. This has a strong impact on how much in the future, as data volumes increase, the cost of enterprise data retention, day-to-day data operations (such as backup), legal obligations, or data-related workloads will increase exponentially over the data lifecycle. This results in that if the cost increases significantly for data that is no longer usable or obsolete, then in practice you will lose data if you do not pay attention to the data lifecycle, at every stage of the lifecycle.
The data lifecycle is considered to begin with the collection or creation of data. This means that when, for example, IoT sensors produce data or new information is entered into information systems. In many of these situations, it is already known whether this transaction will generate new customer information or even billing information that has retention requirements under either data protection or accounting laws, so they should be automatically tagged in the metadata so that they are known in the next lifecycle. In addition to this, data classification and quality control should be done at this stage but more on these topics in our other blog posts.
Once new data is created or collected, it is stored, which is the second stage of the life cycle. At this point, the metadata created in the previous section makes it easier to save in the right place, under the right conditions, and for the right time. It is clear that if the information is not classified, then it can be stored in a less secure data warehouse, which can be significantly less costly. Or, alternatively, if information is rarely needed, it is pointless to store it in a high-availability data warehouse.
The next step in the lifecycle is to use the data, which is significantly easier that the data can be trusted, verified, available, and it is known when the data was collected and whether it is still current. In the use phase, analyzes and the generation of understanding also influence the generation of data value and the consideration of potential other data users. These two factors already provide guidance for the next step in the life cycle, which is data exchange and selling. If your data is one that you do not intend to sell or share, you can proceed directly to the last two steps, namely archiving and finally deleting the data.
Overall, the data lifecycle described above is therefore considerably initial in terms of execution and managing, so it is worth thinking about and implementing the first steps correctly to make each step even easier after that. Thus, there are six stages in the life cycle, each of which can be technically and operationally divided into the necessary sub-stages, but mainly in the following main stages you will survive when you have defined them correctly.
Life cycle in stages:
- Collect or create information
- Save data for use
- Use and analyze data
- Share, exchange and sell information
- Archive data for the end of the life cycle
- Delete data
Please make sure that your management team and your technicians are both aware of each of these stages.