Intelligence · Module
Forecast & Simulation
Forecasts with ranges, backtests and stated assumptions, not a spreadsheet nobody trusts
Forecast & Simulation is planned to forecast certified metrics with P10/P50/P90 ranges, backtest every model against simple benchmarks and build driver-based scenarios with an assumption register. Monte Carlo, sensitivity and stress tests show how likely a goal is.
Planning people can trust
Planning often lives in a spreadsheet with unstated assumptions and single-point numbers. Forecast & Simulation, planned for W4 and requiring BI Studio, is designed to make ranges, assumptions and model quality visible.
Forecasts that prove themselves
The forecasting service works on certified metrics and produces P10/P50/P90 intervals for each horizon. Every model runs rolling-origin backtests with WAPE or MAPE and interval coverage, compared with naive and seasonal naive benchmarks. A model that does not beat seasonal naive is not used. Live forecast error is monitored and the model status is downgraded automatically when error exceeds tolerance. A person must sign off before a model is used in strategic decisions. Calendar regressors include tenant holiday and festival calendars, fiscal periods and promotions.
Scenarios built on drivers
The scenario builder works on BI Studio metric trees, so drivers such as price, conversion, headcount or lead volume flow through to outcomes. Each assumption is recorded with its value or distribution, source, confidence and owner; a run is blocked when an assumption lacks a source or confidence. Monte Carlo runs produce outcome distributions, and scenario comparison shows side-by-side outcomes, the probability of hitting a goal target and the key drivers. Sensitivity (tornado) analysis and stress tests, such as demand down 30% or a two-week supplier delay, are planned.
Connected to the rest of One3
Domain packs arrive as their apps do: pipeline and revenue from CRM, 13-week cash from Finance, ticket volume from Service Desk, demand by SKU and location from Inventory, and workforce capacity from People (advisory only). Scenarios can be published to Goals, FP&A and Decision Engine options, and exported to or imported from Sheets. Discrete-event simulation for queues in support, warehouses and field service is planned for W4 to W5.
Features
What Forecast & Simulation includes
Forecasts with intervals
P10/P50/P90 per horizon for certified metrics.
Rolling-origin backtests
WAPE, MAPE and interval coverage against naive and seasonal naive.
Benchmark rule
Models that do not beat seasonal naive are not used.
Driver-based scenarios
Built on BI Studio metric trees.
Assumption register
Value or distribution, source, confidence and owner per assumption.
Monte Carlo runs
Outcome distributions from labelled input distributions.
Scenario comparison
Side-by-side outcomes and probability of hitting a goal.
Sensitivity and stress tests
Tornado charts and named stress scenarios.
Drift monitoring
Automatic downgrade when live error exceeds tolerance.
Human sign-off
Required before a model is used in strategic decisions.
Calendar regressors
Holiday and festival calendars, fiscal periods and promotions.
Domain packs
Pipeline, 13-week cash, ticket volume, SKU demand and advisory workforce capacity.
Publish and Sheets exchange
Scenarios to Goals, FP&A and Decision Engine; export and import with Sheets.
Workflow
Forecast to approved scenario
- 1
Choose a certified metric and horizon
- 2
Candidate models are backtested against benchmarks
- 3
A planner reviews accuracy and signs off
- 4
Drivers and assumptions are set in a scenario
- 5
Monte Carlo and sensitivity runs show the range
- 6
The chosen scenario is published to Goals, FP&A or the Decision Engine
AI
AI skills and Digital Employees
Forecasting
Statistical and ML forecasts with intervals; strategic use needs sign-off.
Backtesting and drift
Evaluates models and downgrades them automatically when error grows.
Scenario generation
Proposes scenarios from goals and signals for planner review.
Assumption completeness check
Blocks runs with unsourced assumptions.
Digital Employees that use this module
They act through the same governed Actions as people; risky steps wait for approval.
Frequently asked questions
What does it require?
BI Studio, because drivers and forecasts use certified metrics. It is planned for W4.
How do I know a forecast is any good?
Every model is backtested and compared with naive benchmarks; models that do not beat seasonal naive are not used.
Can it plan headcount?
Workforce capacity scenarios are planned as advisory only. People decide on employment matters.
See Forecast & Simulation in your workflow
Tell us about your team and the outcomes you want. We will walk you through the apps, the Digital Employees and the plan for your industry.