There is a lot more to data analysis and insights than simply counting a few vanity metrics over time and placing them on a dashboard.
To begin with, you need to think carefully about the questions that you need to answer. Without this direction it is easy to end up with an output that is quite interesting, but of little value to your business. Once the questions are clear, you should consider the scale of the data: is there enough of it to be useful, or is it currently in a location where it is problematic to process? It is also important to consider data protection and security. You may have sensitive customer data that you need to protect or the data that would be commercially valuable to a third party.
Data projects can become costly and not return any value if approached in the wrong way, which is why we typically conduct a data feasibility study. These studies broadly follow our Discovery framework, but aren’t as in depth across all areas. A feasibility study reduces risk by establishing the current possibilities and highlighting action that need to be taken, such as collecting new data, or making infrastructure changes to enable data transfer.
Once we have established what is feasible and what is going to be of most value to your business, we can begin to deliver your solution. This could include big data processing using R or creating predictive models with Azure Machine Learning and Stream Analytics. The result could be anything from an advanced reporting suite, to a simple dashboard that surfaces the single most important piece of information that you need to act on at any one time.
While the outputs will vary, our goal with data is always the same: to make it valuable.