Explain a Forecast Change Without Hiding It in the New Total
Prepared with AI assistance by Alactic Editorial. Examples identified as hypothetical are illustrative, not customer results.
A forecast can rise while the underlying pipeline weakens. New opportunities might offset lost deals, or a higher assumed win probability might conceal opportunities slipping into the next quarter. The new total alone cannot distinguish those explanations.
To explain sales forecast changes, reconcile two frozen snapshots at opportunity level. Start with the previous forecast, apply additions, removals, amount changes, probability changes, and period shifts in a declared order, and finish at the current forecast. Each movement needs both a calculated contribution and a reason for the changed assumption.
The bridge below uses a probability-weighted forecast. It is a recommended operating design, not a claim about a particular software feature. Its purpose is to make the movement inspectable: another person should be able to reproduce the new total from the old one.
1. Define exactly what the two totals include
Before calculating movement, write a short forecast definition. Specify the target period, amount basis, currency basis, eligible records, and treatment of won and lost business. Otherwise, an apparent commercial change may simply be a change in measurement.
For the worked example, the hypothetical forecast represents expected bookings for one target quarter. Open opportunities contribute amount multiplied by probability. Won opportunities remain in the quarter at 100%; lost or cancelled opportunities contribute zero. A close date outside the target quarter gives the opportunity a zero contribution to that quarter, even if it remains active.
Use the same definition at both snapshot dates. Keep probability conventions stable too: if stage determines probability, document the stage-to-probability mapping. A mapping change is a forecasting-method change, not necessarily new evidence about buyers.
The contribution formula is:
Quarter contribution = Amount × Probability × Period indicator Period indicator = 1 when assigned to the target quarter Period indicator = 0 otherwise
This is not a universal forecast model. If your forecast is an unweighted commit list, use its actual inclusion rule instead. Do not apply probability weighting merely to make this example fit. If management adds an override above the opportunity forecast, reconcile that override separately rather than silently distributing it across deals.
2. Preserve the evidence needed to reconstruct movement
A current export cannot reliably explain what changed. Preserve both snapshots, including stable opportunity IDs, amounts, probabilities, close dates, statuses, and the forecast rules used. Names are useful for readers but unreliable as matching keys.
Include records outside the target quarter when they could move into it. An opportunity that existed previously in a later quarter is a timing shift when it moves in, not a newly created opportunity. Comparing only the two in-quarter lists loses that distinction.
For every changed opportunity, retain:
- Previous and current values: enough to recompute both contributions.
- Movement classification: new, removed, or retained, followed by the applicable retained-record changes.
- Reason: what changed in the commercial assumption.
- Evidence and owner: the supporting observation and the person responsible for the assumption.
An opportunity absent from the current file is not automatically a commercial removal. It might have been deleted, merged, excluded by a filter, or omitted from the export. Resolve those cases before assigning a lost-deal explanation. If an ID changed during a migration, preserve a crosswalk rather than treating the replacement as new pipeline.
3. Use one calculation order to allocate interactions
When both amount and probability change, the two effects interact. There is no unique allocation unless you choose a convention. A useful convention is to change amount first, probability second, and period last for retained opportunities.
- Additions: add genuinely new opportunities at their current quarter contribution. Do not decompose them again.
- Removals: subtract the previous quarter contribution of confirmed exits, such as cancelled opportunities. Exclude them from subsequent calculations.
- Amount changes: for retained opportunities, change amount while holding previous probability and previous period membership fixed.
- Probability changes: use current amount, but hold previous period membership fixed.
- Period shifts: apply the change in period membership using current amount and probability.
For each retained opportunity: Amount effect = (A1 − A0) × P0 × I0 Probability effect = A1 × (P1 − P0) × I0 Period effect = A1 × P1 × (I1 − I0) 0 = previous snapshot; 1 = current snapshot A = amount; P = probability; I = period indicator
These three effects sum exactly to the retained opportunity's contribution change. New opportunities and confirmed removals are handled separately, so they cannot also appear in the amount or probability rows.
The order affects attribution, not the final total. Under this convention, a retained opportunity moving into the quarter contributes its full current weighted value to timing: its earlier amount and probability effects are zero because its previous period indicator was zero. A departing opportunity can have amount and probability changes before its remaining current value is removed through timing. Keep this convention visible so readers do not mistake it for a uniquely correct causal decomposition.
4. Filled fictional example: a forecast rises from $200,000 to $219,000
All opportunities, figures, and explanations in this example are hypothetical. Amounts below are in thousands of dollars. The snapshots are one week apart and refer to the same target quarter.
Opportunity Previous snapshot Current snapshot Atlas 100 × 50%, in quarter 120 × 60%, in quarter Beacon 80 × 75%, in quarter Cancelled; contribution 0 Cedar 60 × 50%, in quarter 60 × 50%, next quarter Delta 40 × 50%, in quarter 50 × 40%, in quarter Elm 100 × 40%, in quarter 100 × 40%, in quarter Fjord Did not yet exist 90 × 50%, in quarter Grove 70 × 60%, next quarter 70 × 60%, in quarter Previous quarter forecast: 50 + 60 + 30 + 20 + 40 = 200 Current quarter forecast: 72 + 0 + 0 + 20 + 40 + 45 + 42 = 219
The filled bridge separates the sources of that $19,000 increase:
Forecast-change bridge ($000) Previous forecast 200 Additions Fjord: 90 × 50% +45 Subtotal 245 Removals Beacon: subtract previous 80 × 75% −60 Subtotal 185 Amount changes, at previous probability Atlas: (120 − 100) × 50% +10 Delta: (50 − 40) × 50% +5 Subtotal 200 Probability changes, at current amount Atlas: 120 × (60% − 50%) +12 Delta: 50 × (40% − 50%) −5 Subtotal 207 Period shifts, at current weighted value Cedar: moves out, −60 × 50% −30 Grove: moves in, +70 × 60% +42 Current forecast 219 Unexplained residual 0
The category totals are additions of $45,000, removals of $60,000, amount increases of $15,000, probability increases net of decreases of $7,000, and timing inflows net of outflows of $12,000. Together they explain the $19,000 net increase.
Do not stop at the net timing figure. Grove brought $42,000 into the quarter while Cedar pushed $30,000 out. Those are different assumptions, even though they partially offset in the bridge.
5. Show why simultaneous changes are not counted twice
Atlas moved from $50,000 to $72,000 of weighted contribution, a $22,000 increase. Its amount rose from $100,000 to $120,000, and its probability rose from 50% to 60%.
The amount-first method assigns $10,000 to amount and $12,000 to probability. The probability row uses the new $120,000 amount, so it already includes the interaction between the larger deal and higher confidence.
Atlas, hypothetical figures: Amount effect: 20,000 × 50% = 10,000 Probability effect: 120,000 × 10% = 12,000 Total change: 22,000 Interaction embedded in probability effect: 20,000 × 10% = 2,000
Calculating both changes using old values would explain only $20,000 and leave the $2,000 interaction missing. Calculating both using new values would explain $24,000 and count the interaction twice.
You could instead report the interaction as a separate category. That makes the algebra explicit but adds another line for managers to interpret. For a practical operating bridge, assigning the interaction consistently to the second calculation is often easier to maintain.
Delta demonstrates another reason to retain both rows. Its weighted contribution stayed at $20,000, but a larger amount was exactly offset by lower probability. A report that lists only opportunities with changed weighted totals would hide both assumption changes.
6. Attach reasons to the assumptions, not just the arithmetic
A reconciled bridge explains where the movement sits. It does not establish whether the revised assumptions are justified. Add a short reason register alongside the calculation.
The following entries are also fictional:
- Atlas, amount +$10,000: revised scope adds a second deployment phase. The opportunity owner should verify the amount against the latest proposed scope.
- Atlas, probability +$12,000: the buyer has confirmed the evaluation criteria. The owner should explain why that evidence supports the probability increase; a positive conversation alone is not a calculation rule.
- Beacon, removal −$60,000: the buyer cancelled the initiative. Preserve the cancellation reason rather than deleting the historical record.
- Delta, amount +$5,000 and probability −$5,000: the proposed scope expanded, but budget availability became less certain. These are separate changes despite no net forecast movement.
- Cedar, timing −$30,000: the expected decision moved beyond quarter-end. Check the revised decision milestone supporting the new close date.
- Grove, timing +$42,000: the expected decision moved into the quarter. Verify the milestone that makes the earlier date credible.
- Fjord, addition +$45,000: newly identified opportunity. Check that it is genuinely new rather than a duplicate or renamed record.
The resulting explanation is more useful than “forecast increased by $19,000.” In this example, new pipeline and incoming timing outweighed a cancellation and outgoing timing, with a further net increase from revised amounts and confidence. Managers can then challenge the specific assumptions carrying the increase.
7. Repair reconciliation failures before interpreting them
Calculate an explicit residual:
Residual = Current forecast
− (Previous forecast + all attributed movements)A nonzero residual means the bridge is incomplete or inconsistent. Do not bury it in “other” and describe the result as explained. Label it unresolved until its cause is identified.
- Different snapshot scope: rerun both extracts with the same eligibility rules and check record counts.
- Missing or duplicated IDs: repair matching before interpreting additions and removals.
- Won deals disappearing: if the forecast includes booked business, retain won contribution at 100%. Otherwise a win can misleadingly look like a forecast loss.
- Currency or model changes: hold the conversion basis constant for commercial attribution, or show a separate explicit adjustment. Apply the same discipline to changed probability mappings.
- Manual overrides: bridge the opportunity subtotal and the override separately, then reconcile the combined forecast.
- Rounding: calculate using full precision and round for display. Identify any display-only rounding difference.
Record corrections also need honest labels. Fixing a mistaken amount changes the forecast, but it does not represent newly won scope. Preserve the arithmetic category while marking the reason as a correction.
8. Make the next comparison reproducible
For the next forecast update, freeze the two snapshots, apply the calculation order above, and fill one bridge for the target period. Keep the detailed opportunity rows beneath the summary so every category total can be traced to records.
Use three observable completion checks: the unexplained residual is zero at calculation precision; every changed contribution has an attribution; and every material changed assumption has an owner and supporting reason. Define “material” before reviewing the movements, and retain smaller changes in the calculation even if their explanations are grouped.
Finally, inspect opportunities like Delta whose net contribution did not move. The next action is not merely to explain the $19,000 increase. It is to identify which amount, probability, and timing assumptions now support the forecast—and which still need verification.
Put the next conversation to work.
Explore Alactic Communications for WhatsApp conversations, campaigns and AI-assisted workflows. Choose the product and plan that fit your team.
Explore Communications