Pandoras Analytics is focusing on predictive analytics. Before being able to predict the future, however, data flows through several stages of the data analytics value chain, which ultimately leads to operational and strategic business decisions:
Information is the first step into analysis to find out what happened. It’s where you separate signal from noise. Algorithms can help to find patterns in data and together with domain expertise you can identify relevant patterns. Since information describes what happened in the past, it is called descriptive analytics.
Forecasting is the foray into predictive analytics to identify what it all means for the future. There are many advanced analytics methods available, but the most critical part is to understand that all projections are wrong. Business modeling techniques are then used to identify a range of potential outcomes instead of focusing on only one. Forecasting is as much art as it is science. Domain expertise is combined with quantitative analytics to model the future.
After all the analyses, it’s time for business decisions. These can be operational decisions impacting the next few weeks, months, or quarters and strategic choices impacting the next few years. Although business know-how is critical here, it’s imperative to understand the prior stages including how the data was clustered, what happened in the past and why, and what are the inherent limitations of the forecasting and modeling techniques.