Back to Pricing Dashboard User Documentation - Main Page
The aim of the Forecast Evolution page is to:
- Make historical comparisons: by comparing past forecasts to the current live forecast, users can understand how accurate previous forecasts were and identify any trends or patterns in forecasting errors.
- Evaluate performance: By comparing the frozen forecast with actual results, users can evaluate the performance of their forecasting models and make necessary adjustments to improve accuracy.
Data accessibility is linked to the visibility model, if your profile does not allow to see costs/ICM data, you will not see all the information of this page!
Forward Evolution Primary Rules
- Today, DL keeps 2 forecasts per day starting from 8th day of the month.
- To limit data volume in Pricing Dashboard, the first snapshot of the week is prioritized, i.e., users will have 4 snapshots in 1 month.
- As every snapshot starts with the ongoing month at the time of its creation (ex: March 2025 for the version of March 2025), it becomes necessary to complete their past horizon to get a proper comparison.
- To do so, actual sales are replicated to fill the missing months (until the beginning of the year) as follows:
Forward Evolution Navigation
Select a Start and End Run Period ie the 2 forecast periods that you wish to compare.
- Version selection: Users shall select both forecast run periods to be compared.
- Unit selection: Users can choose which currency and unit to use.
- Time filters: Users can filter on time dimensions to reduce his/her scope of comparison.
- Dimension selection: Users can change the time dimension used for evolution charts.
- Quick filters: Users can use predefined filters on time dimensions.
- KPI selection: Users can either focus on sales amounts or quantities.
- Forecast Evolution over Time: This chart provides a global comparison for 2 run periods over time, comparing old forecasts to new actuals in the process.
- Forecast Period: This chart provides a high level comparison of multiple run periods, following how the overall amounts were estimated more or less optimistically week after week.
The next section shows the reasons behind the variation:
- Dimension selection: Users select which dimension to use in the related chart(s) to explore their figures.
- Display selection: In regards to the left dashboard, users select whether they display those gaps as values (amounts or quantities) or percentages.
- KPI selection: Users can either focus on sales amounts or quantities.
- Gap Overview: This chart provides a global comparison (at high level) of both selected run periods, either as amounts or percentages over the related dimension value.
- Top/Bottom charts: Those charts offer the capability to explore the differences at a deeper level, highlighting the biggest/lowest gaps.
Detailed table allows the user to see the variation at a granular level, dimensions based on those defined by the GBU:
- Activation: To avoid any performance issue, users have to use this button for activating/deactivating the table, and then preventing any calculation of oversized scopes of data.
- Detail selection: Users can choose either to focus on a more or less detailed table.
- Scale selection: Users can choose to level up the scale of the figures to avoid displaying huge unreadable numbers.
- Detailed table: The table provides a (more or less) detailed view of the gaps observed.