Running Forecast Calls

• Source: Report research library
• Frameworks applied: ORM-Tech forecast categories, Rework deal inspection evidence bands, Rework four-tier pipeline review cadence, Signals 60-minute CRO forecast call, Clari required-coverage (1 / win rate)

Category entry criteria are evidence tests, close-date changes force demotion, and probabilities are explicitly banned from categories.

Known weakness, published as found: Coverage math (1/win rate, stale-pipeline haircuts) duplicates forecasting-pipeline-coverage; reduce to the number needed for the call and link out. Add one filled example row to the top-10 inspection table showing what an acceptable 'next proving event' looks like.

1. 1.Copy the SKILL.md text below, or download the raw file.
2. 2.Create a folder named exactly running-forecast-calls in your agent's skills directory.
3. 3.Save the file inside that folder as SKILL.md.
4. 4.Ask the agent one of the trigger requests below.
5. 5.Check the output against what you already know before it leaves your desk.

• Our commit keeps slipping every quarter — help me rebuild how we run the weekly forecast call
• Here is our open pipeline export; tell me which of these deals actually belong in commit and which should be demoted
• Draft a 60-minute forecast review agenda for my sales managers with the questions I should ask on each deal

• Build the ARR bridge and efficiency metrics slides for next week's board meeting
• Design next year's AE quotas and commission accelerators for the mid-market team

---
name: running-forecast-calls
description: >-
  Runs a disciplined weekly revenue forecast call: applies hard evidence-based
  entry criteria to Pipeline / Best Case / Commit / Closed Won, inspects deals
  against probability bands, enforces a fixed 60-minute agenda, and emits a
  forecast call packet with a committed number and dated action items. Use when
  the user says "run our forecast call", "forecast review agenda", "clean up our
  commit", "why do we keep missing the forecast", "our commit slipped again",
  "pipeline review agenda", "deal inspection", "what should be in commit", "is
  this deal really a commit", or asks how to structure a weekly CRO forecast
  meeting. Use it whenever the task is judging whether specific deals belong in
  a forecast category, even if the user does not say "forecast call". Do NOT use
  for building a board revenue package (see building-board-revenue-reporting),
  quota or commission design (see designing-sales-compensation), or a new
  leader's onboarding plan (see planning-cro-first-90-days).
metadata:
  version: "1.0"
---

# Running forecast calls

Run one weekly forecast call: re-categorize every in-period deal against evidence tests, inspect the deals that carry the number, and leave with a committed figure plus owned actions. Building the board package, redesigning quotas, and long-range capacity planning are out of scope.

Note on sourcing: the forecast-category criteria, inspection bands, and call agenda below come from RevOps consultancies and vendor resource libraries, not from primary academic research. The mechanics are well specified and usable; treat the accompanying rates as directional vendor figures.

## Workflow

Copy this checklist into your reply and tick items as you complete them:

```
- [ ] 1. Pull the in-period dataset and freeze it
- [ ] 2. Re-categorize every deal against entry criteria
- [ ] 3. Assign an evidence band to each Commit and Best Case deal
- [ ] 4. Compute coverage from the real win rate
- [ ] 5. Validate the packet against the hygiene gate; fix and re-run
- [ ] 6. Run the 60-minute agenda
- [ ] 7. Ship action items within 30 minutes
```

**1. Pull and freeze the dataset.** Export every open opportunity with a close date in the period plus everything closed to date, carrying these fields: amount, close date, close-date change count and last change date, stage, current forecast category, last activity date, named economic buyer, named champion, next step and its date, and competitor. Freeze this export as the call's version of truth. Do not re-query mid-call, because the roll-up will move under the discussion and the number on the board becomes unreproducible.

**2. Re-categorize every deal.** This step is low-freedom: apply the tests literally, do not soften them for a favored deal. All criteria must be true for a category, and a deal falls to the highest category it fully earns ([ORM-Tech](https://orm-tech.com/blog/sales-forecast-categories-explained/)).

| Category | All must be true |
|---|---|
| Closed Won | Countersigned order or PO in hand |
| Commit | Close date in period **and unmoved for two weeks**; economic buyer confirmed; pricing and terms agreed; order form or procurement in motion; no open blocker owned by the customer |
| Best Case | Close date in period; champion confirmed; proposal delivered; **exactly one** named risk with a dated mitigation |
| Pipeline | Qualified; next step booked; meaningful activity in the last 30 days |
| Omitted | Disqualified, pushed out of period, or stale — untouched for 12 months |

Two rules that do the actual work:

- **Any close-date change demotes the deal.** A deal cannot remain in Commit after its date moves; it drops to Best Case or Pipeline until it re-earns Commit criteria ([ORM-Tech](https://orm-tech.com/blog/sales-forecast-categories-explained/)).
- **Never attach probabilities to categories.** Categories are a shared evidence language, not a likelihood scale. Assigning 90/60/30 percent re-imports the rep judgment the criteria exist to replace, and makes a Commit from one team incomparable to a Commit from another ([ORM-Tech](https://orm-tech.com/blog/sales-forecast-categories-explained/)).

**3. Assign an evidence band.** For every Commit and Best Case deal, and for any deal at 3–5× average deal size, assign one of five bands — 80–100%, 60–80%, 40–60%, 20–40%, 0–20% — from verified evidence rather than stage position ([Rework](https://resources.rework.com/libraries/pipeline-management/deal-inspection-process)). Judge the band across six dimensions: opportunity overview and what changed since last review; qualification validation against the record; stakeholder verification classifying every contact as Champion / Supporter / Neutral / Skeptic / Blocker with decision authority and rep access noted; competitive position; timeline reality check; risk identification. This step is high-freedom on the band value and low-freedom on the dimensions: the source publishes the bands but not the scoring rubric that maps evidence to a band, so state your reasoning per deal and keep the mapping consistent across reps rather than pretending to a formula. Inspect on these triggers: a requested stage progression, before any deal enters Commit, any deal at 3–5× average size, any stalled deal, and every forecast deal at the start of a quarter ([Rework](https://resources.rework.com/libraries/pipeline-management/deal-inspection-process)).

**4. Compute coverage from the real win rate.** Use required coverage = 1 ÷ trailing win rate, not a fixed 3× ([Clari](https://www.clari.com/blog/pipeline-coverage-best-practices/)). A 33% win rate needs 3×, 25% needs 4×, 20% needs 5×, and enterprise motions at 15–25% win rates need 4–7×. Applying 3× to an enterprise motion with a 15% win rate can mean missing the number by 40% ([Clari](https://www.clari.com/blog/pipeline-coverage-best-practices/)). Strip stale deals first: a team reporting 4× with 30% stale pipeline actually holds 2.8× qualified coverage, and deals aged beyond 2× the average sales-cycle length should be discounted or removed ([Clari](https://www.clari.com/blog/pipeline-coverage-best-practices/)). Where cycles run 120 days or longer, compute coverage on a rolling two-quarter window.

**5. Validate the packet, then loop.** Run this gate before the call and fix every failure:

```
- Sum of Closed Won + Commit + Best Case reconciles to the stated call number
- Zero Commit deals with a close-date change inside the last 14 days
- Zero Commit deals with no activity in 14+ days
- Zero Commit deals without a named economic buyer
- Every Best Case deal has exactly one named risk with a dated mitigation
- Every Commit and Best Case deal has an evidence band and a next-step date
- Stale deals (12 months untouched) moved to Omitted before coverage is computed
```

Re-run the gate after fixing. Only proceed to the call when it passes clean, because a call that opens by arguing about hygiene never reaches the decisions segment.

**6. Run the 60-minute agenda.** Hold the timings; they are the mechanism that keeps the call from becoming storytelling ([Signals](https://www.signals.ceo/resources/cro-weekly-forecast-call)).

| Minutes | Segment | Content |
|---|---|---|
| 0–5 | Number on the board | Commit, best case, gap to plan. No context, no story. |
| 5–25 | Deal-by-deal | Top 10 by ARR: what changed since last week, and the next event that confirms or kills the deal. |
| 25–40 | Pattern review | Slipped deals, stage regressions, coverage movement. |
| 40–55 | Decisions | Resource deployment, executive sponsorship, which reps need help this week. |
| 55–60 | Action items | Owners and dates. |

Reps present their own top deals; managers roll up the long tail, because rep presentation creates accountability while manager roll-up of the tail is the only way the call fits in 60 minutes ([Signals](https://www.signals.ceo/resources/cro-weekly-forecast-call)).

Ask these three questions on every deal, verbatim ([Signals](https://www.signals.ceo/resources/cro-weekly-forecast-call)):

- "What is the next event that proves this deal will close on the date you committed?"
- "Who at the customer, by name, has agreed to sign?"
- "What did you learn this week that changed your confidence?"

When an answer is vague, drill rather than accept. "They're evaluating the proposal" becomes: what specifically are they evaluating, against what requirements, who is doing the evaluation, when do they finish, and what happens after ([Rework](https://resources.rework.com/libraries/pipeline-management/pipeline-reviews)).

**7. Ship action items within 30 minutes.** Every action records What / Who / When / Why, and the next call opens by reviewing last week's actions before any new deal is discussed ([Rework](https://resources.rework.com/libraries/pipeline-management/pipeline-reviews)). Track Action Item Completion Rate — the percentage completed by due date — as the measure of whether the call is working at all.

## Red flags to call out live

Raise any of these on the deal in question rather than saving them for a summary ([Rework](https://resources.rework.com/libraries/pipeline-management/pipeline-reviews)):

- No activity in 14+ days: not an active opportunity, a hope. Email-only or rep-initiated-only contact counts the same.
- Close date pushed three times, or a close date three weeks out on a deal created last week, or "close by end of quarter" with no stated business driver.
- Budget described as "allocated" with no amount, source, or approval path named.
- Single-threaded champion, or a champion who will not introduce the rep to approvers.
- Incumbent competitor with no articulated reason for the buyer to switch.

## Output format

Emit this exact structure; downstream forecast tracking parses the headings. Wording inside sections is yours.

```
## Forecast call packet — <team> — <period>, week <n>

**Number on the board**
Closed won: $X | Commit: $X | Best case: $X | Plan: $X | Gap to plan: $X
Weighted coverage: N.Nx | Required coverage (1 / win rate NN%): N.Nx

**Category changes this week**
| Deal | Amount | From | To | Trigger |

**Top 10 inspection**
| Deal | Amount | Band | Next proving event | Date | Owner | Risk |

**Patterns**
- Slipped: <count, $, where>
- Stage regressions: <count, where>
- Coverage movement vs last week: <delta and cause>

**Decisions made**
- <decision> — owner — date

**Action items (What / Who / When / Why)**
- <action> | <owner> | <date> | <why>
```

## Gotchas

- Roughly 20% of pipeline dated to close in a quarter, measured on day one of that quarter, actually closes in-quarter — 80% does not arrive on time ([ORM-Tech](https://orm-tech.com/blog/sales-forecast-categories-explained/)). Treat day-one in-period pipeline as a claim, not a base case, and never quote it to Finance as a landing range.
- The strongest single risk signal is also nearly the last one you get. Ranked earliest to latest: no activity or field changes (earliest, strong, usually missed) → champion stops responding → close date changed (late, strongest) → deal slipped while sitting in Commit (definitive, too late). Build the weekly inspection around activity and field staleness, because by the time the close date moves the quarter is already decided ([ORM-Tech](https://orm-tech.com/blog/sales-forecast-categories-explained/)).
- More than 10% of average pipeline has gone untouched for 12 months ([ORM-Tech](https://orm-tech.com/blog/sales-forecast-categories-explained/)). Coverage computed before purging it overstates the quarter, so purge first and report the purged figure as the real number.
- A large gap between average open-pipeline deal size and average closed-won deal size is inflation, not ambition — ORM-Tech illustrates it as $80K open versus $40K won ([ORM-Tech](https://orm-tech.com/blog/sales-forecast-categories-explained/)). Compare the two every week and haircut amounts on deals priced above the closed-won average.
- Reviewing coverage in week 10 of a 13-week quarter leaves no runway to correct ([Clari](https://www.clari.com/blog/pipeline-coverage-best-practices/)). Coverage is a week-1-to-week-4 decision input; after that the only lever left is deal-level execution.
- Deal inspection depth beats breadth: a 45-minute manager 1:1 should inspect 3–4 deals thoroughly at 10–15 minutes each rather than touching 10–15 superficially ([Rework](https://resources.rework.com/libraries/pipeline-management/pipeline-reviews)). If the forecast call is the only inspection happening, the categories will decay regardless of how tightly the call is run.
- Declining forecast accuracy is a review-rigor signal, not a modeling problem. Track forecast accuracy trend, review frequency per rep, and deal velocity after review together; when accuracy slides while activity looks fine, inspection quality has slipped ([Rework](https://resources.rework.com/libraries/pipeline-management/pipeline-reviews)).
- The CRO owns the operating forecast and the actions behind it; the CFO owns financial rigor and how numbers reach board reporting. When the two run different assumptions, the forecast call degrades into a reconciliation exercise ([Weflow](https://www.weflow.ai/blog/cro-30-60-90-day-plan-framework)). Agree the assumption set before the call, not during it.

What does the Running Forecast Calls skill do?

Runs a weekly forecast call with hard category entry criteria, evidence-band deal inspection, and a fixed 60-minute agenda

Where does the Running Forecast Calls skill come from?

Report research library. It was written by The Revenue AI Report against a 12 criterion quality rubric and graded in an independent scoring pass.

Why was the Running Forecast Calls skill chosen for this library?

Category entry criteria are evidence tests, close-date changes force demotion, and probabilities are explicitly banned from categories.

When should the Running Forecast Calls skill not be used?

Do not use it for: Build the ARR bridge and efficiency metrics slides for next week's board meeting Or: Design next year's AE quotas and commission accelerators for the mid-market team

How do I install the Running Forecast Calls SKILL.md file?

Download the file, create a folder named exactly running-forecast-calls inside your agent's skills directory, and save the file inside it as SKILL.md. The agent loads it when a request matches the description.