A rigorous versioning and locking process is essential for running meaningful forecast accuracy reporting and for giving your supply chain team a stable number to plan against. Confido has a full suite of versioning tools designed to support standard S&OP and IBP cycles.
8.1 Scenarios and Versions
Confido supports multiple forecast versions simultaneously. You can create a new version by saving an existing one and giving it a name — for example, “P6 Upside,” “P6 Base,” or “P6 Low Side.” Each version is independent and can use different model configurations or adjustments. Each saved version is also available in the reporting module, allowing you to compare and contrast the differences between versions.
Common uses for multiple versions:
A “live” version that the demand team actively maintains and hands off to supply chain
A “high side” scenario that includes all distribution opportunities at full likelihood, for capacity planning
A “budget” version loaded at the start of the year for ongoing variance tracking
A “sales forecast” version that represents the unadjusted sales team view, for comparison
To save a version, click the “Save Snapshot” button in the top right corner of the demand planning workspace.
Tip: Naming conventions matter. Use a consistent naming structure (e.g., “P6 2026 — Live,” “P6 2026 — High Side”) so that versions are easily identifiable in reports and review meetings. |
8.2 Snapshotting for Mix Forecast Accuracy Tracking
A snapshot is a point-in-time copy of a forecast version that is saved and locked — it cannot be edited after it is created. Snapshots are the foundation of forecast accuracy reporting.
The typical workflow for accuracy tracking:
At the end of each planning cycle (e.g., end of period), Confido automatically takes a snapshot of the live forecast and stores it.
This snapshot becomes your “lagged forecast.” Lag 1 is defined as the forecast that was in place one period before actuals arrived, Lag 2 as the forecast that was in place two periods before actuals arrived and so forth.
When actuals come in, a dedicated report labeled “Mix Forecast Accuracy and Bias” in the reporting section automatically compares actual shipments to each historical snapshot at whatever level of granularity and lag you need (refer to section 8.3).
Confido can also set up automated snapshots on a schedule — for example, automatically taking a snapshot every Friday at 6 PM — so you always have a consistent lock point without relying on manual action.
8.3 Forecast Accuracy and Bias Reporting
Confido’s Omni-powered reporting layer lets you build fully customizable forecast accuracy dashboards. Common accuracy reports include:
Any lag accuracy by customer / product — how accurate was the forecast X periods ago compared to actual shipments that came in? Sortable by MAPE or raw unit variance
Biggest misses — a ranked list of the customer-item combinations with the largest absolute or percentage deviation between forecast and actuals, for the current period or rolling window
Accuracy trend over time — how is forecast accuracy improving or declining over rolling periods, by customer or total business
Model comparison — for customers where you are evaluating different statistical models, how does each model’s historical accuracy compare to the sales forecast?
Reports can be configured to filter by any attribute (customer, channel, item, product family) and can be exported to Excel or scheduled for automatic email delivery to stakeholders who do not have direct Confido access (refer to section 9).