What is the Stagger Chart table?
Accessible from the KPI tab by clicking the Details icon, the Stagger Chart table lets you analyze forecast history and assess forecast quality over time.
Unlike a standard view, which only displays the latest forecast, this table retains all forecasts archived at each period. This makes it possible to compare different versions of the same forecast, measure their evolution, and identify discrepancies between successive forecasts and actual observed demand.
Access and navigation
In the KPI tab, click the Details icon to open the pop-up window containing the table.
The forecast tracking table corresponds to the first icon (1), below the navigation arrows. The interface arrows (2) let you move from one item to another; you can also use the keyboard arrows for faster navigation.
At the top right of the window, the Forecast Details button (3) gives access to the details table in the Forecast Edition tab.
Analysis settings
Two drop-down menus, located at the top of the table, let you customize the analysis:
Lag: the offset between the forecast date and the period being analyzed.
KPI analysis window: the number of periods taken into account when calculating performance indicators (for example, the last 3 periods).
Cells outlined in blue indicate the forecasts corresponding to the selected lag. When you change the lag, this selection automatically shifts within the table to display the forecasts used for calculating the KPIs with the newly chosen lag.
Summary KPIs
Three indicators, displayed below the filters, provide an overview of forecast quality over the selected period:
Average error: the average deviation between forecasts and actual demand.
Average bias: indicates whether forecasts tend to be systematically over- or under-estimated.
Average accuracy: reflects the overall reliability level of the forecasts.
These three KPIs are automatically recalculated based on the chosen lag and analysis window.
Customizing the display
A navigation bar lets you adapt how the table is displayed to suit your needs:
Error, Bias, or Accuracy: choose which KPI to display in the matrix.
% / Units: display results as a percentage or as absolute values.
Display scale (Full / k / M): display values in full, in thousands, or in millions, particularly useful for rounding financial amounts and making them easier to read.
Forecast baseline : overlay the initial forecast generated by SKU Science to compare it with the forecasts adjusted by planners.
Reading the table
For each demand period, the table shows the various forecasts made over time.
Each row corresponds to an archived forecast, i.e. a version saved during an earlier planning cycle. The column headers indicate the dates of past and future periods.
This layout makes it possible to:
view all the historical forecasts kept in the archives;
track how the forecast evolves from one period to the next;
compare successive forecasts with actually observed demand;
analyze the stability of forecasts and the impact of adjustments made during the demand planning process.
This view helps you understand how a forecast was built up over time, measure the improvement — or deterioration — in its quality, and identify opportunities to optimize the forecasting process.
Concrete analysis example
Let's take item 00-1 at lag 1. The tool's base forecast was 1,063 units, and the user adjusted it to 1,000. However, actual demand came to 1,872: the error between this demand and the user's forecast at lag 1 (1,000) is therefore 47%.
When the Forecast Baseline display option is enabled, a color code indicates whether the user's adjustment brought the forecast closer to or further from actual demand: in red, the change worsened the forecast (the gap with actual demand widened); in green, it improved the forecast (the gap narrowed).
The example below also makes it possible to track how a forecast evolves from one cycle to the next. This is the purpose of the right-hand part of the table, titled Forecast Horizon: the percentages shown there indicate this evolution. For example, the forecast for the month of July recorded in April was 800 units; having risen to 1,000 units in May, it shows a change of +25%.
Key takeaways
The Stagger Chart table offers a unique overview of how a forecast is built and evolves across planning cycles. By retaining the full history of archived versions, it not only makes it possible to measure the accuracy and bias of past forecasts, but also to assess whether the adjustments made by planners actually improved or worsened forecast quality. It is a valuable tool for identifying friction points in the forecasting process and targeting areas for improvement, in support of more reliable and higher-performing demand planning.