Sales forecasting in Dynamics 365

Forecasting is where sales-management discipline gets tested. Dynamics 365 Sales ships a configurable forecasting module that fits the cadences most B2B sales organisations actually run.

The model

A forecast configuration defines the dimensions of the forecast: the sales hierarchy (org chart, territory, product line), the period (monthly, quarterly), and the measures to surface for each cell of the grid. Default measures: Submitted (the seller's committed number), Best case, Pipeline, Won, Quota. Custom measures (rollups, calculated columns) can be added.

Hierarchy

A forecast is most useful when it rolls up. The standard rollup is the manager hierarchy (your team aggregates into you), but territory-based and product-line-based hierarchies are equally common — and configurable on the same data. Multiple forecasts per organisation are normal.

Predicted forecast

Beyond seller-submitted numbers, Dynamics 365 Sales runs an AI-predicted forecast based on historical close rates, opportunity attributes, stage progress, and engagement signals. The predicted forecast appears alongside the submitted forecast in the grid, giving managers a number to compare against.

Snapshots

Forecasts are versioned and snapshotted — daily by default — so you can see how the forecast for next quarter has drifted week over week. The snapshot history is the basis for forecast accuracy analysis and for spotting drift early.

Manager adjustments

A manager can override an aggregated number with a manager judgment. Both the original rolled-up value and the override are stored, so audit and accuracy analysis can see where overrides happened and how they panned out.

Workflow

Forecasts have submission states — Draft, In Review, Submitted — and can be locked at a snapshot deadline so the team commits to a number at a moment in time.

Inline editing

Sellers and managers work directly in the forecast grid; clicking an opportunity opens a slide-out for in-place editing without leaving the forecast.

Copilot

Generative AI explains the forecast: why is this quarter trending down, which deals are slipping, which late-stage opportunities lack required actions. Natural-language queries return data without the seller leaving the grid.

Integration

Forecasts can publish to Power BI for executive dashboards, and to Excel for finance integration. Many companies still take the final number to a separate budget/forecast process in finance.

Operational reality

The mechanics are easy; the discipline is hard. Forecasting only works if pipeline hygiene is real — stale opportunities, optimistic close dates, and dead leads pollute the forecast as much as the AI predictions can correct.