Back to all articlesForecasting discipline

Account for Deals That Can Slip Together

Alactic Editorial9 min read

Prepared with AI assistance by Alactic Editorial. Examples identified as hypothetical are illustrative, not customer results.

Three opportunities have different owners, different buyers, and separate close dates. Yet all three require the same parent company to release a budget. If that decision moves beyond quarter-end, the opportunities may move together—even if each salesperson has done everything expected.

To account for correlated opportunities in sales forecasting, identify the shared decision, connect every affected opportunity to it, and model outcomes for the group. Keep the weighted forecast, but add joint scenarios showing what happens when the common dependency clears or slips.

The important distinction is mathematical as well as operational: dependence can change the distribution of outcomes without changing any individual opportunity’s expected value. A reasonable weighted total can therefore hide an unreasonable assumption about how much downside is diversified.

1. Find the decision behind the close dates

A close date describes when the seller expects an outcome. A shared dependency describes something that must happen for several outcomes to remain possible within that period.

Look for a common decision-maker, source of funds, purchasing window, or external approval. Opportunities can share one of these even when they sit under different accounts or belong to different sales teams.

  • Shared budget release: several subsidiaries are waiting for the same corporate allocation.
  • Shared purchasing window: separate departments must submit purchases during one centrally managed buying round.
  • Shared external approval: multiple buyers depend on one funding or programme decision before they can authorise purchases.

Do not group deals merely because they belong to the same industry or have similar close dates. Those similarities can prompt investigation, but they do not establish a common gate. The useful question is: What identifiable event could move these opportunities together?

Keep this exercise focused on forecast uncertainty. A shared implementation resource or delivery schedule is a different planning problem unless it directly changes whether the purchase can close in the forecast period.

2. Build a dependency map with evidence and a fallback

A useful map needs more than a tag such as “budget risk.” It should name the decision, show the affected amount, and explain what happens if the decision misses its required date.

Use one record per shared dependency, with linked opportunity references. This is a recommended operating design; a shared worksheet is enough to begin. The key fields are:

  • The dependency identifier and the specific event being awaited.
  • The opportunities and amounts exposed to that event.
  • The latest decision date consistent with an in-period close.
  • The evidence supporting the connection and when it was checked.
  • A named owner responsible for checking the shared event.
  • Any verified alternative funding or purchasing route.

Distinguish the latest workable decision date from the customer’s hoped-for decision date. A budget release just before quarter-end may be too late if purchase authorisation still requires additional steps. Establish that timing from the buyer’s process rather than assuming the release itself equals a closed deal.

Fictional worked example: one budget decision, three opportunities

All names, amounts, dates, probabilities, and outcomes below are hypothetical. The example uses thousands of a single reporting currency, written as “k.” Amounts represent the organisation’s chosen bookings measure, not recognised revenue.

DEPENDENCY MAP — HYPOTHETICAL QUARTER-END REVIEW

Dependency: BUD-01
Event: Northstar Group releases its discretionary purchase budget
Latest workable release date: 12 September
Forecast period ends: 30 September
Dependency owner: Group account lead
Evidence: Central purchasing contact says the three purchases
          require this release; separate buyer confirmations recorded
Last checked: 3 September
Alternative route: None verified
If late: Move affected purchases outside the current quarter
Next check: 8 September, with central purchasing contact

Opportunity       Amount   Required event   Current weight
Northstar East     100k    BUD-01 clears       60%
Northstar West      80k    BUD-01 clears       60%
Northstar Central   60k    BUD-01 clears       60%
Harbor renewal      50k    Separate decision   80%

Gross amount linked to BUD-01: 240k
Weighted amount linked to BUD-01: 144k
Harbor weighted amount: 40k
Total weighted forecast: 184k

The map makes two different exposures visible. There is 240k of gross opportunity amount attached to one decision, and 144k of expected contribution attached to it. Neither figure means that 144k will actually close.

The Harbor renewal provides a contrasting opportunity: in this hypothetical model, its outcome is independent of the Northstar budget decision. That independence is an explicit assumption, not something established by its different account name.

3. Preserve the expected value, change the joint model

For a deliberately simple first model, assume the budget release is the only remaining uncertainty for the three Northstar opportunities. If it arrives in time, all three close this quarter. If it does not, none closes this quarter. Assign the release a hypothetical 60% probability.

This is a strong assumption, useful for exposing the shared risk. It is not a claim that budget approval normally guarantees a sale. A more realistic version can retain additional uncertainty after the release.

Under this simplified model, each Northstar opportunity still has a 60% individual close probability:

Northstar East:    100k × 60% = 60k
Northstar West:     80k × 60% = 48k
Northstar Central:  60k × 60% = 36k
Harbor renewal:    50k × 80% = 40k

Expected total = 184k

Expected values add whether or not the opportunities are independent. So the weighted total is not mathematically wrong simply because the deals are correlated.

What changes is the range and likelihood of joint outcomes. If the three Northstar deals were independent at 60% each, the probability that all three missed would be 40% × 40% × 40%, or 6.4%. In this shared-gate model, all three miss whenever the budget slips: a 40% probability.

Similarly, the hypothetical chance of all three closing rises from 21.6% under independence to 60% under the shared gate. The model puts more probability at both extremes. It does not merely reduce the forecast.

4. Put the joint scenarios beside the weighted number

The following filled scenario table combines the shared budget outcome with the separate Harbor renewal. Because Harbor is assumed independent of the release, each joint probability is the product of the two relevant probabilities.

JOINT SCENARIO TABLE — ALL FIGURES HYPOTHETICAL

Budget release   Harbor   Probability   Northstar   Harbor   Total
On time          Closes      48%          240k       50k     290k
On time          Slips       12%          240k        0k     240k
Late             Closes      32%            0k       50k      50k
Late             Slips        8%            0k        0k       0k

Probability check: 48% + 12% + 32% + 8% = 100%
Expected total: (48% × 290k) + (12% × 240k)
              + (32% × 50k) + (8% × 0k) = 184k

The expected total remains 184k, but no scenario produces 184k. More importantly, there is a hypothetical 40% probability of finishing at 50k or less.

If the operating plan requires at least 150k, the model assigns a 40% probability of falling short. That threshold is a hypothetical management requirement, not a recommended universal target.

The decision-ready forecast statement is therefore not just “weighted pipeline is 184k.” It is: “Expected bookings are 184k, but 240k of gross pipeline depends on one budget release. Under the stated assumptions, a late release leaves at most 50k this quarter.”

This separates the centre of the forecast from its downside. It also gives managers a concrete event to monitor instead of asking several owners for nearly identical confidence updates.

5. Relax the all-or-nothing assumption when evidence supports it

The first table is intentionally stark. Real groups may have partial funding, different purchasing thresholds, or unresolved deal-specific issues. Refine the model where those differences matter.

For each opportunity, ask two conditional questions: What is its chance of closing this period if the shared event happens on time? What is its chance if the event is late?

For example, an alternative hypothetical version of Northstar East might assign an 80% chance of closing after an on-time release and a 30% chance after a late release because a verified smaller funding route remains available. With a 60% chance of timely release, its overall close probability would still be 60%: (60% × 80%) + (40% × 30%).

That example preserves the individual expected value while weakening the all-or-nothing link. Other opportunities need their own conditional assumptions. Do not copy one pair of probabilities across the group merely to preserve existing weights.

Also distinguish a conditional expected subtotal from an actual scenario outcome. If the “budget on time” row contains probability-weighted deals, that row is a conditional mean, not a promise that the subtotal will occur. Label it accordingly. Otherwise, the revised model can hide uncertainty inside the very table intended to reveal it.

6. Avoid fixes that make the forecast less trustworthy

Do not apply an unexplained correlation discount. Reducing all affected weights and then adding another group-level haircut can count the same risk twice. State whether a weight already includes the budget uncertainty. Change individual probabilities when evidence changes; change the joint model when the relationship between outcomes changes.

Do not treat every shared account as one block. A centrally funded purchase and a locally funded renewal may move differently. If a buyer confirms that one opportunity has independent authority and available funds, remove or weaken its link rather than preserving a tidy group.

Do not mistake missing information for independence. An unconfirmed funding source should remain an open question. Record who will verify it and by when. Absence of a documented link is not evidence that no link exists.

Do not manufacture precise probabilities. If the release probability is poorly supported, show alternative cases. For this hypothetical example, the expected total is 136k at a 40% release probability and 232k at an 80% release probability, keeping Harbor’s expected contribution at 40k. That sensitivity may be more useful than defending an unsupported 60% estimate.

Start with dependencies large enough to change a planning decision. Mapping every weak relationship creates maintenance work and can distract from the one gate that dominates the period.

7. Turn the map into a review and recovery routine

In the forecast review, discuss the shared event once, then review genuine differences between the linked deals. The dependency owner should bring the latest evidence; opportunity owners should explain remaining buyer-specific risks and any verified exceptions.

If the decision misses its latest workable date, update the group together. Preserve the original forecast snapshot, record the changed evidence, and revise the in-period scenarios. Distinguish “slipped out of period” from “lost”; a timing failure is not automatically a lost sale.

Recovery should address what can actually change. Verify whether an alternative funding route exists, whether some purchases can be independently authorised, and whether unaffected pipeline provides credible coverage. Do not replace the slipped group with unrelated opportunities simply by increasing their weights.

For the next review, take the largest opportunities in the current period and ask their owners to identify the funding authority, purchasing window, and external decision gate. Produce one dependency map and one scenario table for the largest confirmed shared exposure.

Measure the exercise by whether it reveals a planning-relevant concentration: the gross amount linked to the gate, its weighted contribution, and the probability of missing your chosen operating threshold. Then track whether the shared event happened when expected and which linked opportunities actually moved together. Those observable outcomes tell you whether the dependency model needs strengthening, splitting, or retiring.

Sales forecastingPipeline managementScenario planningRevenue operations

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