Generate reports

Generate reports

You can generate different reports which are based on article attribute values and targets. To do so, click Generate Reports, the following window appears:

You can deselect any of the reports, if you want then click Generate, the following reports are generated:

Architecture Summary

This report appears as shown in the below example:

This report is calculated based on the following logic:

FW19FW20ChangeTRGT%TT
STYLESInputted from TargetsCount from Line Builder – total rows(FW20-FW19)/FW19Inputted from Targets(FW20-TRGT)/TRGT
ARTICLESInputted from TargetsSum from Line Builder – total of “Total Articles”(FW20-FW19)/FW19Inputted from Targets(FW20-TRGT)/TRGT
PRODUCTIVITYInputted from TargetsWHSL REV/ARTICLES(FW20-FW19)/FW19Inputted from Targets(FW20-TRGT)/TRGT
WHSL REVInputted from TargetsSum of “GMP Grade $ (Whsl)”(FW20-FW19)/FW19Inputted from Targets(FW20-TRGT)/TRGT
MARGINInputted from Targets(Sum of “Margin $ Ext”)/(Sum of “GMP Grade $ (Whsl)”(FW20-FW19)/FW19Inputted from Targets(FW20-TRGT)/TRGT

Note:

  • Values in this report are changed according to the data filtered using the Filter tab.

Style Intro Flow

This report appears as shown in the below example:

This report is calculated based on the following logic:

  • NEW: Count from Line Builder where “Style N/CO” = “NEW” and “Intro Month” = “Jan”
  • C/O: Count from Line Builder where “Style N/CO” = “C/O” and “Intro Month” = “Jan”

The logic is the same for all the months.

Article Lifecycle

This report appears as shown in the below example:

This report is calculated based on the following logic:

  • 6 Month Articles: Sum from Line Builder – the total of “6 Month Articles”
  • Spring: Sum from Line Builder – the total of “Spring Articles”
  • Summer: Sum from Line Builder – the total of “Summer Articles”
  • Prefall: Sum from Line Builder – the total of “Pre-Fall Articles”

Product Ranking, Platform, and Subsilhouette Distribution

This report appears as shown in the below example:

This report is calculated based on the following logic:
Product ranking for platform #1 and SubSilhouette #1:

  • Best: Count of rows within Line Builder that fit criteria (Best, Platform #1, Class #1)
  • Better: Count of rows within Line Builder that fit criteria (Better, Platform #1, Class #1)
  • Good: Count of rows within Line Builder that fit criteria (Good, Platform #1, Class #1)

Product Segmentation

This report appears as shown in the below example:

This report is calculated based on the following logic:
Product ranking:

  • Best: performance is Count of rows within Line Builder that fit criteria (Best, Performance)
  • Better: performance is Count of rows within Line Builder that fit criteria (Better, Performance)
  • Good: performance is Count of rows within Line Builder that fit criteria (Good, Performance)

Product Ranking, Platform and Segment Distribution

This report appears as shown in the below example:

This report is calculated based on the following logic:
Product ranking for platform #1 under customer segmentation #1:

  • Best: Count of rows within Line Builder that fit criteria (Best, Platform #1, Segment #1)
  • Better: Count of rows within Line Builder that fit criteria (Better, Platform #1, Segment #1)
  • Good: Count of rows within Line Builder that fit criteria (Good, Platform #1, Segment #1)