When a sales forecast rises or falls, the new total does not explain the movement. Leaders need to know whether the change came from new opportunities, deals moving between periods, wins and losses, or changes in deal value and probability. A forecast bridge turns two dated CRM snapshots into a deal-level reconciliation.
This guide focuses on one practical view: the unweighted amount of open opportunities expected to close in a selected period. A weighted forecast or a manager-submitted forecast is a different measure and needs its own bridge. Keeping those views separate prevents a pipeline total from being mistaken for a forecast call, bookings, or revenue.
For a related metric, see how to measure sales forecast accuracy. Accuracy evaluates a frozen forecast against what happened later; a bridge explains how the forecast changed between two points in time.
Choose exactly what the bridge measures
Before exporting CRM data, write down the measure, period, filters, and cutoff. “Forecast” can refer to several numbers, and they do not move in the same way. Use the same definition at the beginning and end of the bridge.
| Measure | What is counted | Question it answers |
|---|---|---|
| Open pipeline amount | Unweighted amount of open opportunities that meet the period and scope rules | How did the amount of open deals in this period change? |
| Weighted expected value | Opportunity amount multiplied by the chosen probability or weight | How did the modeled value change under the stated weighting? |
| Submitted forecast | A saved rep, manager, or leadership forecast under the organization’s definitions | How did the forecast judgment or category change? |
| Bookings or recognized revenue | Closed commercial outcomes or accounting revenue under their respective rules | What was booked or recognized? These are outcomes, not open pipeline. |
The example later in this article uses open pipeline. A won deal leaves that open-pipeline total, so its amount appears as a negative movement in this specific bridge. If your measure is a forecast that includes booked deals, the treatment changes. State the rule on the report instead of silently applying a pipeline convention to another number.
Freeze two comparable snapshots
A bridge needs a starting state and an ending state for the same close period. Save the actual data extracts or system snapshots with their “as of” timestamps. Record the time zone, currency, pipeline, team or segment filters, open/closed definition, and whether the view is amount-based or weighted. Set an as-of date before comparing reports so a late edit does not change the baseline unnoticed.
At minimum, retain a stable opportunity ID, owner, status, stage, close date, amount, currency, probability or forecast category if used, and the timestamp at which the state was captured. Add the event or field history that supports any change label. Salesforce, for example, documents pipeline-change and forecast charts, as well as opportunity history; the available views depend on product features, field tracking, permissions, and configuration. A current CRM screen alone may not reconstruct a past state.
If historical data was not retained, do not backfill a supposed prior forecast from today’s values. Start storing snapshots now, disclose the missing baseline, and use the first comparable pair only after both are available. Salesforce describes its own historical reports and snapshots in its documentation; other CRM systems have different capabilities and retention rules.
Classify each deal once
For an open-pipeline bridge, compare each opportunity’s prior and current state and assign one primary movement bucket. Apply the rules in a fixed order so a deal that was won, for example, is not also counted as pushed out or repriced.
| Movement | Deal-level rule | Direction in this open-pipeline example |
|---|---|---|
| New in period | Not in the prior snapshot; present and in scope at the current snapshot | Add the current eligible amount |
| Pulled into period | Previously outside the selected close period; now inside and still open | Add the current eligible amount |
| Pushed beyond period | Previously inside the period; now expected to close after it and still open | Subtract the prior eligible amount |
| Won | Previously open and in scope; now closed-won | Remove from open pipeline; show bookings in a separate outcome view |
| Lost | Previously open and in scope; now closed-lost | Remove from open pipeline |
| Amount change | Open and in scope at both snapshots; deal amount changed | Current amount minus prior amount |
| Scope or data correction | Merge, duplicate, currency, eligibility, or filter correction changes inclusion | Show as a separate adjustment with evidence |
New and pulled-in are different. A new opportunity did not exist in the baseline record set. A pulled-in opportunity existed but was outside the selected period. If a record is missing from the old export and its history cannot show why, put it in an unresolved or data-quality line rather than claiming it is new demand.
Changes to stage or forecast category do not change an unweighted amount by themselves. They matter in a weighted or submitted-forecast view. For a weighted expected value, each deal’s value is amount multiplied by its selected probability. If amount and probability both change, one transparent way to split the combined movement is:
Amount effect = (current amount − prior amount) × (prior probability + current probability) ÷ 2
Probability effect = (current probability − prior probability) × (prior amount + current amount) ÷ 2
These two effects add exactly to the change in amount × probability. They are an accounting decomposition, not proof that a stage change caused a buyer to become more likely to purchase. Probability definitions and forecast categories are organization-specific; document the version used at both snapshots.
Build the reconciliation from record history
- Lock the scope. Set the same forecast period, currency basis, team, status rules, and metric for both snapshots.
- Match stable IDs. Join the two states by opportunity ID, not company name or a row number that can change.
- Apply the movement rules. Give each opportunity one primary transition. Keep merge, deletion, and scope corrections visible rather than hiding them inside “other.”
- Calculate the delta. For each movement bucket, use the value appropriate to the defined metric. Sum each deal once.
- Tie the bridge out. Confirm that the start value plus all signed movements equals the end value. Investigate any unexplained remainder before presenting the result.
Salesforce’s Pipeline Inspection documentation describes a waterfall with movement measures such as opportunities created, increases or decreases, wins and losses, and amounts moved into or out of a period. Its forecast charts also describe changes over time. These are useful platform examples; a team should verify the fields, filters, and history its own system actually retains.
Worked example: the open pipeline rose by $250,000
Suppose the open, unweighted opportunity amount expected to close this quarter was $1.20 million on October 1. On October 8, the same view totals $1.45 million. A synthetic reconciliation is:
| Movement | Change | Why it belongs here |
|---|---|---|
| New opportunities in the quarter | +$300,000 | Not present in the baseline; now open and in the selected close period |
| Pulled into the quarter | +$200,000 | Previously expected after quarter-end; now expected in the quarter |
| Pushed beyond quarter-end | −$150,000 | Previously in the quarter; now expected later |
| Won and removed from open pipeline | −$100,000 | Closed-won deals are no longer open pipeline |
| Lost and removed from open pipeline | −$50,000 | Closed-lost deals are no longer open pipeline |
| Amount changes on deals still in scope | +$50,000 | Current eligible amounts exceed the baseline amounts |
| Net movement | +$250,000 | $300,000 + $200,000 − $150,000 − $100,000 − $50,000 + $50,000 |
The tie-out is $1.20 million + $0.25 million = $1.45 million. If the numbers do not reconcile, inspect duplicates, currency conversion, missing records, period boundaries, edits between extracts, and changes to filters. Do not force the bridge to balance by adding an unexplained plug.
The example is synthetic and measures open pipeline only. It does not say that the new opportunities were caused by a particular campaign, that the pushed deals will close later, or that the final quarter’s bookings will be $1.45 million.
Read the movement without overclaiming
A bridge answers “what changed in this defined number?” It does not, by itself, answer why the buyer changed course. Follow up on the few movements that materially affect the decision: inspect the opportunity history, confirm the buyer-side evidence, and ask the owner what changed. Treat the reason as unknown when the CRM only shows a date or stage edit.
For example, a large pulled-in amount may reflect a real buyer-confirmed milestone, a rep’s optimistic date, or a cleanup before the forecast meeting. The bridge reports the movement; the underlying record and buyer evidence inform the interpretation. Keep this work distinct from a close-date-only analysis and from the weekly pipeline review agenda.
When the goal is to diagnose whether pipeline was created or existing deals simply progressed, use a separate view such as pipeline creation versus progression. A useful bridge can complement that analysis without combining unlike stages into one headline.
Pre-presentation checklist
- Both snapshot timestamps and the selected close period are visible.
- The measure is named: open amount, weighted value, or submitted forecast.
- Currency, team, pipeline, and inclusion rules match across snapshots.
- Every opportunity maps to one movement bucket, or to a documented unresolved line.
- New and pulled-in deals are not mixed.
- Won/lost treatment matches the selected measure.
- Amount or probability effects are separate when both can move.
- Start plus signed movements equals end; the remainder is zero or explained.
- Large changes link to record history and a named owner; buyer explanations are evidence-based.
- Public examples use synthetic or safely aggregated information, not identifiable CRM records.
Frequently asked questions
What is a sales forecast bridge?
It is a reconciliation that explains the difference between a defined starting forecast or pipeline value and a comparable ending value by listing the movements that account for the change. A deal-level bridge makes each movement traceable to an opportunity and its history.
What data is needed to build one?
Use two dated opportunity snapshots with stable IDs and the fields needed for the selected measure, such as status, stage, amount, close date, probability, and forecast category. Add field or stage history when available. If neither history nor an earlier snapshot exists, the missing baseline cannot be reliably recreated from current values.
How is a bridge different from forecast accuracy?
Forecast accuracy compares a forecast frozen at a known time with a later outcome. A forecast bridge compares two forecast states to explain how the estimate changed. A team may use both: the bridge for movement diagnosis and accuracy for retrospective scoring under a stated formula.
How often should the bridge be refreshed?
Refresh it when leaders need to explain a decision-relevant change and the source system retains comparable history. There is no universal weekly or monthly rule. Match the cadence to the sales cycle, forecast review, and quality of available snapshots; preserve the same cutoff logic so periods can be compared.
Sources and further reading
- Salesforce Help: Pipeline Inspection Charts — platform-specific waterfall and movement-chart behavior.
- Salesforce Help: Pipeline Forecasting Charts — week-to-week changes and comparisons with prior forecast periods.
- Salesforce Help: Opportunity History — field and stage history behavior.
- Salesforce Help: Report on Historical Changes — historical trend reports and reporting snapshots.
- Salesforce Help: Why Opportunity Trends and History Reports Differ — explains the distinction between those data sources.
How did this article land?
Choose one reaction. You can change it anytime.