Types of Information and Databases

At a high level, we can divide information into one of two types: transactional or analytical. Transactional information refers to the granular data generated by real-time business processes. Real-time means that the data are recorded temporally as the transaction occurs. For example, consider what data might be generated by an ATM transaction (date and time, amount, customer, account, location), a customer checkout (items, quantities, payment method, date, clerk, customer), or an airline reservation (flight number, seat number, payment method, customer info). Transactional data are stored in what we call an operational database. Operational databases are those required for everyday business processes or operations. Therefore, it is critical that operational databases remain running and available at full speed. Otherwise, the business shuts down. Speed and availability are the key performance indicators of an operational database storing transactional data.

Analytical information refers to any data or information that can be used to make intelligent inferences relevant to the organization, help solve unstructured business problems, and support higher-level, ad-hoc decision making. While transactional information simply helps you perform your business processes, analytical information is used to help you gain strategic competitive advantage in the marketplace. Before reading any further, ask yourself, "what information would be useful for these purposes and where would it come from?"

Answer: Whereas transactional information only includes the data that are recorded as part of our own organization's processes, analytical information could come from both inside and outside the organization. For example, summarized transactional information would be analytical information. Therefore, analytical information includes all of the transactional information (although usually summarized by product, region, customer, etc.). However, it may also include competitor information scraped from the web. It could also include paying for access to third party databases like those provided by the major credit bureaus. Perhaps it would include hiring a consulting firm to research new customers or survey your existing customers for satisfaction and loyalty data. Or, if your organization has a mobile application, then it may include the location data collected from customer cell phones. The point is, you have to be creative and intelligent when it comes to analytical information.

Lastly, analytical information is not stored in operational databases. Remember that speed and availability are the key characteristics of operational databases. So, if managers want to pull out some data to analyze or add some new data they bought from a third party, then they should not mess with the operational databases, potentially slowing down or crashing the business. Therefore, to be safe, we copy all of the transactional data from operational databases, transform the data into a summarized form, and load everything into an analytical database (a.k.a. data warehouse) where managers can analyze the data to their heart's content. We have a lot to learn about data warehouses and data analytics. However, that will come later. For now, you simply need to know that transactional and analytical information are stored separately.