There is a major change happening in the IT industry — the use of big data and analytics to guide how businesses are run. Many companies are embracing analytics as part of their core strategies. Unfortunately, some of those companies think that they can just purchase an analytics solution, implement it, load and collect data, and poof, all their problems will be solved! There are many software solutions available for business analytics, but there is also still a need for requirements analysis to ensure the right solution is selected and implemented correctly.
Business analytics implementations need software requirements, just like any other commercial off the shelf (COTS) solution. There will be business requirements, functional requirements, non-functional requirements, and business rules. Requirements models help us to understand how people will use the solution and to guide the configuration of the system according to the business rules.
The process to elicit and development requirements for business analytics projects can be broken down into four steps as shown in the following diagram. In this blog post we’ll go into detail on the fourth step, defining the data transformation analysis requirements. More information can be found on the other three steps from our whitepaper, “Forward Thinking for Tomorrow’s Projects: Requirements for Business Analytics.”
In an analytics project, once you have identified the data and understand how users want to consume and use the results of analytics solutions, then the main focus is on determining what analyses must convert the data into get those results.
There are some situations in which the users just want to look at the raw data objects and attributes to explore them, because they don’t really know what they want. However, particularly with big data, it’s important to at least zero in a bit on what kinds of problems they want to solve and therefore what type of analysis they might want on the data — otherwise the solution might be too overwhelming to be useful. Good business analysts (BAs) can brainstorm about problems to help stakeholders figure out what they want from the solution. In fact, BAs can do quite a bit of research to determine what possibilities exist with analytics solutions to guide the brainstorming process.
Business analytics solutions can enable future-state strategic analysis, such as exploring “what-if” scenarios. So, BAs can help by posing “what if” scenarios that might be of interest to the stakeholders. For example, “what if we could offer a new service and predict what our future sales would be for that service, before we ever deploy it? How would that be helpful?” A good analytics solution with the right data can run models and algorithms to enable these types of data predictions. That said, prediction models and usage scenarios still must be specified in the requirements so that the analytics system can be configured correctly. Furthermore, the analyses that transforms the data might require computations, statistics, or other business rules to be applied to the data prior to it being delivered in the solution.
