With nextanalytics, we have taken a different approach to analytics. Our goal is to enable a junior or part-time business analyst to quickly provide insightful, ad-hoc information analysis to mainstream business managers.

You will not see a lot of statistic formulas or complicated concepts --  most business analysis relies on simple math; differences, growth, percentages, subtotals, average. When we do dip into a more advanced transformation such as a Frequency Distribution, we reduce it to a simple command (GetCountsOfValues) that most people will be able to understand.

Approach: Many analytics references suggest that you should start with your business goals, develop the questions, determine the data needed and then start the analysis. This is usually a good approach to follow, especially if you are dealing with typical business intelligence tools that require a lot of setup and configuration effort. With nextanalytics, there is almost no setup required, so we will assume you are starting with some data and a general idea of the questions and business goals, and that your analysis is ad-hoc and make change as you explore the data in various ways.

Orientation: nextanalytics holds your data in ‘pages’ of rows and columns, similar to database tables or spreadsheet worksheets, and operates on them using script commands. Most script operations create a new page to hold the results of the operation, with the original page kept in memory unchanged. A number of commands can reference multiple pages to do things like filtering or comparisons. You can also load data from different sources, each loaded into its own page, and have them all available in memory to be operated on by a script command.

As an example, the picture shows a newsletter campaign analysis that starts by loading a page of target values as well as a page of metrics for each of the newsletters. The metrics page is split into two pages, calculations are performed and the results pulled back into a common page. Some additional calculations are then performed to determine the alert colors, and the results page is then colored with alert colors indicating how well the newletters did with respect to targetted values for each of the metrics.