With the release of the Extensible Analytics Engine, GoodData has improved on existing capabilities in developing ranking metrics.
GoodData’s ranking functionality has been improved with the following new features:
- Ranking functions for use within MAQL metric definitions
- Dense rankings
- Percentage ranks
- Cumulative distribution
- Ranking within report sub-groups
You can apply a RANK function to any GoodData metric using MAQL in the Custom Metric Editor. Suppose you are using one of the basic metrics in a GoodData Sales Analytics project: Avg. Won. This metric represents the average size of a won opportunity.
The report below uses the Avg. Won metric and slices it by a Quarter/Year (Snapshot) attribute.
Original single-metric report
For simplicity’s sake we’ve filtered for the years 2011 and 2012, to limit our results to just eight rows of data.
Now suppose you want to rank all the results of the Avg. Won metric on the level of Quarter/Year. To do so, we’d add a second metric to the report:
Below, the results display ranking for all rows from smallest to largest values. By default, ranking is carried out in ascending order (keyword
ASC), but you can also specify descending order ranking by adding the DESC keyword after the metric specification:
Report with a ranking metric added
If you were to change the report granularity to
month/year, the ranking function would rank the new results accordingly – adapting to the new granularity of hte report.
However, if you want to rank the metric results within subgroups in the report, such as by
Year, while continuing to view data at the
Quarter/Year dimensionality, we’ll need to add a
WITHIN(Year) statement to our metric definition. Our new metric would take the following form:
Ranking within report subgroups
For more on RANK function syntax, including additional variations of the WITHIN statement, see our article Introducing the Extensible Analytics Engine.
And as always, for a comprehensive overview of MAQL, see MAQL - Multi-Dimension Analytical Query Language.