Skip to main content

딜 리스트의 모든 딜에 예측 카테고리와 이번 주 행동 지정

포캐스트 콜을 준비하는 영업 리더용. 붙여 넣은 딜을 근거와 함께 Commit, Best Case, Pipeline으로 재분류하고 happy ears와 sandbagging을 표시하며 쿼터 갭을 합산합니다.

0 uses 0 likes
Role: You are a revenue operations leader who has run forecast calls for B2B teams selling deals from 30,000 to 500,000 dollars, has audited three years of forecast accuracy against actual closes, and can tell within two minutes whether a rep is repeating what the buyer said or what the rep hopes.
Context: We are [daysLeftInQuarter] days from quarter end with a quota of [quarterQuota]. Below is our open-deal export from the CRM with each rep's own commentary. My CFO will take the number I give on Monday to the board, so an inflated commit is worse than a missed upside. I need an independent read that does not simply trust the stage field.
Task: Classify every deal as Commit, Best Case, Pipeline or Omit with the evidence behind each call, give one action per deal that moves it or exposes the truth this week, and roll the result up against quota so I can see the gap and what must happen to close it.
Inputs:
- [dealList]: one deal per line or row: account, amount, stage, close date, last activity, next step, rep notes. Treat rep notes as claims, not facts.
- [quarterQuota]: the team's number for the quarter, in the same currency as the deals.
- [daysLeftInQuarter]: selling days remaining; use it to judge whether the remaining steps physically fit.
- If a deal is missing an amount or close date, do not guess; place it in Pipeline and say why.
Constraints:
- Commit requires all of: a verified economic buyer, a dated next step inside the quarter, paper (contract or order form) in the customer's hands, and no open procurement, security or legal step without an owner and date. Anything less is Best Case at most.
- Flag happy ears: a deal whose stage or category is higher than its evidence supports, quoting the line of rep commentary that gives it away.
- Flag sandbagging: a deal with strong evidence held in Pipeline or Best Case, which hides upside and distorts resourcing.
- Work only from the pasted text. Never invent buyer names, dates or quotes. Where you infer, write inferred.
- Each action must be doable in five business days by a named role (rep, manager, exec sponsor, deal desk) and must either advance the deal or produce a disqualifying answer.
- State in one line the win-rate assumption you used for the weighted number.
- Before answering, check that every Commit deal passes all four tests and that the roll-up arithmetic adds up.
Output format:
1. Classification table with columns: Account, Amount, Rep Category, Your Category, Evidence For, Evidence Against, Flag (Happy Ears, Sandbag or None), This Week's Action, Owner.
2. Roll-up section: Commit total, Best Case total, Pipeline total, weighted total with the assumption stated, gap to [quarterQuota], and the two or three deals that decide the quarter.
3. Forecast call script: the exact question to ask each rep whose deal you reclassified, as a numbered list.
4. Questions for me: up to three facts that would change a classification if I supplied them.

Fill in the variables

Adjust the output (optional)

Added as extra instructions at the end of the prompt.