The tool can be used to compare compare scenarios or to compare periods.
PERIOD AND SCENARIO COMPARISON
An example of scenario comparison is Actuals vs Plan variance, while an example of period comparison is 2019 vs 2018.
The period you want to compare might be directly mapped in a column. In this case the column - a "Period" column in our parlance - might have two values - 2019 and 2018 you want to compare.
Alternatively, if you have a column in data format - "a Date" column in our parlance - you can choose the time aggregation the app must your to aggregate the periods you want to compare.
If you supply the tool with a date column, the tool will allow you to:
- compare the calendar years against each other
- compare equivalent periods of 12 months
- compare quarter, months or weeks
- compare period-to-date values
The "Compare period to date" widget is set on True by default. The app will compare the appropriate "period to date" (year-to-date, quarter-to-date or month-to-date) periods based of the selected date aggregation period.
Therefore, for example, if you have 2017 data up to the 20th of November, by default the tool will compare January 2017 to November 20, 2017 versus January 2016 to November 20, 2016.
If you set the Date aggregation widget to "Quarter" it will compare "Quarter to date" periods. The forth quarter starts the 1st of October. So it will compare the period between the 1st of October and the 20th of November 2017 to the period between the 1st of October and the 20th of November 2016. If you set the Date aggregation widget to "Month" it will compare the period between the 1st of November and the 20th of November 2017 to the period between the 1st of November and the 20th of November 2016.
If your selected aggregation period is not "Year" the compare "Compare with year before" widget will appear. If set on True, the app will compare the chosen period with the corresponding period of the previous year.
If data aggregation is set to year, it will compare to 12 month periods: for example, if you have 2017 data up to October, by default the tool will compare November-2016 to October 2017 versus November-2015 to October 2016.
If data aggregation is set to Quarter, it will compare the chosen quarter to the corresponding quarter of the preceding year. If the compare period to date widget is set to True it will compare quarters made up of the same number of days. If the widget is set to False, it will compare the full quarter of the previous year to the available data (whether complete or not) of the quarter of this year.
Given this parameter setting....
...if you open the "Dataset info" expander the tool will indicate the most recent date in the dataset, from which it is "starting back" 12 months, which in this case is the 15th of March 2015.
The chart total labels are set to "Total amount Year N-1" and "Total amount Most_Recent_12_Months." Since we are comparing the instances of measured data (not actual to plan) the previous period bar is colored in light grey, while the actual data bar is colored in black, as per the IBCS standard recommendations.
If there is not enough data to go "backwards" by 24 months from the most recent date the app will return a warning...
...and fall back to a simple calendar year comparison.
You can choose the reference period you want to select for the comparison with the "most recent period" widget. If you choose "n" it will select the most recent period in the dataset as the "second" period in the variance analysis comparison.
If you choose "N-1"....
... it will choose the second most recent periods as the reference period and show the new end date of the reference period.
The chart total label and the variance values will naturally change.
COMPARING QUARTERS, MONTHS OR WEEKS
The tool can also compare quarters, months or weeks. If you select month and set "compare with year before" to True it will compare with the corresponding month of the year before. Otherwise it will compare with the preceding month.
All charts will plot the chosen date aggregation comparison.
INVERTING PERIOD ORDER
You can invert the order of the comparison, and set the most recent period as the "first" period in the comparison.
If you click on "dataset info" you will get short report with the periods found in the dataset.
The shown periods will changed depending on the date aggregation choice. Here we chose "quarter".
If you have multiple periods in your dataset - 2014, 2015, 2016 and 2017 - and you want to understand what happened across the whole timeframe, year after year the tool can help you in this analysis by letting you run the three different analyses - 2015 vs 2014, 2016 vs 2015 and 2017 vs 2016 - very easily. However, you will have to stitch the three results together by yourself, out of the system.