Forecasts¶
Forecasts are predicted future values derived from driver trees, formulas, and scenario assumptions. They set expectations for stakeholders and provide the baseline against which actuals are compared.
Who this is for¶
End User Executive Admin
Prerequisites¶
- A ValuePact account with identified value drivers
- At least one value metric with a baseline
- A configured value pack with formulas
Overview¶
Forecasts in ValuePact are scenario-driven and evidence-backed. They are not static spreadsheets; they update dynamically when inputs, baselines, or formulas change.
Forecast methods¶
The platform supports two primary methods:
| Method | Description | Best for |
|---|---|---|
| Formula-driven | Baseline + formula + assumptions = forecast | Quantitative drivers with known inputs |
| Benchmark-informed | Industry peer data informs the target range | New initiatives without historical baseline |
Formula-driven forecasts are computed by the Layer 3 formula engine. Benchmark-informed forecasts pull percentile ranges from Layer 6 datasets.
Scenario planning¶
Every forecast is computed across three scenarios:
| Scenario | Assumption adjustment | Use case |
|---|---|---|
| Conservative | 80% of expected benefit, 120% of expected cost | Risk-averse planning, board reporting |
| Expected | Baseline assumptions from value model | Standard planning, most likely case |
| Optimistic | 120% of expected benefit, 90% of expected cost | Best-case planning, upside communication |
To configure scenario assumptions:
- Open the account's Value Studio workspace.
- Click Forecasts.
- Select a driver.
- Edit the scenario multipliers or input overrides.
- Click Recalculate.
Confidence intervals¶
Forecasts display confidence intervals based on:
- Input variance — standard deviation of historical actuals (if available)
- Model confidence — derived from evidence quality and source count
- Benchmark spread — p25 to p75 range from Layer 6 peer data
A forecast with high evidence quality and low input variance shows a narrow confidence band. A forecast with limited data shows a wide band and is flagged for review.
Time horizons¶
Forecasts can span:
- Short-term — 3 to 6 months, tied to project milestones
- Medium-term — 1 to 2 years, tied to initiative targets
- Long-term — 3 to 5 years, tied to strategic planning
The default time horizon is pulled from the value pack. Users can override it per account.
Forecast revision history¶
Forecasts are versioned. Each revision captures:
- Revision number — auto-incremented
- Author — who made the change
- Change summary — auto-generated diff of variables and assumptions
- Timestamp — when the revision was saved
To view history:
- Open Value Studio > Forecasts.
- Click Revision History next to the driver name.
- Select a revision to preview or restore.
Restoring a revision creates a new revision with the old values, preserving the audit trail.
Locking forecasts¶
Once a business case is approved, its underlying forecast is automatically locked. Locked forecasts display a padlock icon. To modify a locked forecast, you must create an amendment or a new forecast branch.
Example¶
A forecast for "Labor Efficiency" across three scenarios:
| Scenario | Annual Savings | Confidence Interval | NPV (3yr) |
|---|---|---|---|
| Conservative | $480K | $420K–$540K | $1.1M |
| Expected | $600K | $540K–$660K | $1.4M |
| Optimistic | $720K | $660K–$780K | $1.7M |
The business case is built on the Expected scenario, with the Conservative scenario used for risk disclosure.
Permissions required¶
| Role | Permission | Scope |
|---|---|---|
| Admin | Configure forecast templates and scenarios | Organization |
| User | Create and edit forecasts | Assigned accounts |
| Executive | View forecast comparisons across accounts | Organization |
forecasts:write to edit; forecasts:read to view.
Limits and guardrails¶
Limit A maximum of 20 forecast revisions per account per month.
Limit Forecast time horizons cannot exceed 5 years.
Limit Confidence intervals are hidden if fewer than 3 historical data points exist for the metric.
Troubleshooting¶
Issue: Forecast does not update after changing a variable
Cause: The variable is not linked to the forecast driver, or the forecast is locked. Resolution: Verify the variable mapping in Value Studio > Driver Tree. If the forecast is locked, create a revision.
Issue: Confidence interval is extremely wide
Cause: The metric has high historical variance or low evidence quality. Resolution: Add more evidence sources, tighten the baseline measurement, or reduce the scenario multiplier range.
Issue: Benchmark-informed forecast shows 'no data'
Cause: The account industry does not match a Layer 6 benchmark dataset. Resolution: Verify the account industry in Account Settings, or switch the forecast method to formula-driven.
Issue: Restored revision does not update linked business case
Cause: Business cases reference forecast snapshots, not live forecasts. Resolution: Regenerate the business case from the restored forecast revision and submit it for re-approval.
Related pages¶
Escalation path¶
For scenario configuration issues, contact your value pack admin. For forecast computation errors, open a support ticket with the account ID and driver name.