Descriptive analytics is about what happened.
All organizations use descriptive analytics in some form. Common outputs of descriptive analysis include pre-configured dashboards, Excel charts, and sometimes ad hoc exploration of data with statistical software. Successful descriptive analytics help stakeholders understand the present state of business, and then ideate and identify opportunities for change. Once a proposed action is identified, the organization would proceed to use predictive analytics to help the organization understand the possible impact of a proposed action.
- An analyst, at a grocery chain, might have a sales performance dashboard and notices the ice cream category is doing worse than last year. They double click and identify a brand and a flavor that appears to be doing poorly. The analyst identified an opportunity for better assortment planning.
- Another analyst, at a retail bank, might observe that total checking balances held at the bank missed targets. Diving deeper, they notice that new customer originations, specifically from the Northeast have decreased. The analyst identified an opportunity to improve acquisition campaigns.
As a critical root of ideation and opportunity identification, it is important not to neglect investments in good descriptive analysis in favor of more trendy concepts. Descriptive analytics are conceptually simple, there are still significant challenges unlocking the full value of descriptive analytics.
- Data science teams are frequently tasked with ad hoc cuts since other teams may not be SQL or programming savvy. BI tools and dashboards are easy to use, but are most helpful when describing a macro situation or reporting on higher level results.
- The world has become more data enriched, and so the size of datasets continue to grow. Processing time takes longer, making it hard to interactively work with data or to readily get to the cuts that we want.
- It is not always easy to determine what descriptive factors to consider, among all the data dimensions that exist. Guiding users to more interesting insights leads to more hypothesis generation.
Einblick was built to solve these challenges, and empower intuitive interactions with data. Our platform marries the simplicity of no-code output creation with advanced progressive sampling that allows for realtime interaction with very large datasets. Learn more today.