LearnGrokUse one scorecard to review an active opportunity against the qualification criteria your team has already agreed. It gives account executives a defensible next step and gives sales managers a consistent basis for pipeline review.
This method works best when you complete it before a forecast call, deal review, or proposal approval. It does not replace seller judgement. It makes the evidence behind that judgement visible.
Key point
Score evidence, not optimism
A high score needs a specific source, date and owner. A seller's belief is a gap until a buyer confirms it.
1. Set the qualification criteria once
Start with your existing sales methodology, stage exit criteria, or manager checklist. Do not ask the model to invent a qualification framework for each deal. The point is to compare opportunities using the same standard.
Create a scorecard template with six to eight criteria. For each criterion, define what counts as strong, partial, and absent evidence. A practical set is:
| Criterion | Strong evidence | Weak or missing evidence |
|---|---|---|
| Business problem | Buyer described a specific problem and consequence | General interest or seller assumption |
| Priority and timing | Buyer gave a target date and reason for it | Unconfirmed quarter or no event driving action |
| Decision process | Steps, approvers and decision date are known | “They will decide soon” |
| Stakeholders | Champion, economic buyer and users are identified | One contact only |
| Commercial fit | Scope and buying route are discussed | Price sent without buying context |
| Competition and status quo | Alternatives and doing nothing are understood | Seller has not asked |
| Next meeting | Dated meeting has a purpose and buyer attendees | “Follow up next week” |
Use a three-point scale: 2 for evidenced, 1 for partial, and 0 for absent or contradicted. Add a short definition beside each criterion in the template. This stops two managers applying different standards to the same evidence.
Note
Keep the threshold visible
Decide in advance what outcome each total means. For example, a deal with an absent decision process or no next meeting may be paused regardless of its total score.
2. Assemble the deal pack
Gather the material before you ask for an assessment. Use current records, not your memory of the account.
Include:
- The opportunity name, amount band, stage and target close period.
- Notes from the last two customer conversations, with dates.
- Emails that confirm priorities, attendees, timing or next steps.
- CRM fields for contacts, close plan and opportunity history.
- The current proposal or scope, if one exists.
- Known competitors, incumbent supplier and internal alternatives.
- Your qualification criteria and scoring definitions.
Remove personal data that is not needed for the review. Keep the source and date for each item. If a call note says “budget confirmed”, but the note has no date, customer wording or named contact, mark it as unverified rather than treating it as fact.
Check the applicable product and data handling guidance before you paste customer material. Product behaviour and available controls can vary, so use the current xAI documentation overview as your starting point.
3. Request a structured scorecard
Give the model the criteria first, then the deal pack. Tell it not to fill gaps with plausible assumptions. Ask for an output you can paste into the CRM or bring to a manager review.
Use this prompt structure, replacing the text in square brackets:
Assess this active sales opportunity against the qualification criteria below.
Rules:
- Use only the evidence provided.
- Quote or closely cite the source and date for every score of 2.
- Mark unknown information as “not evidenced”. Do not infer it.
- Separate customer facts, seller claims and open questions.
- Identify contradictions between sources.
- Recommend pursue, pause or disqualify. Explain the decision in three sentences or fewer.
Qualification criteria and scoring:
[Paste each criterion and its 0, 1 and 2 definitions]
Deal pack:
[Paste opportunity details, call notes, emails, CRM fields and proposal details]
Return this format:
1. Overall score: [total]/[maximum]
2. Decision: pursue, pause or disqualify
3. Scorecard table: criterion, score, evidence, evidence source/date, gap or risk
4. Stakeholder map: name, role, influence, stance, evidence, missing contact
5. Risks and contradictions, ranked high, medium or low
6. Next actions: action, owner, customer-facing purpose, due date, proof of completion
7. Five questions for the next customer conversation
Ask for a stakeholder map even when the CRM lists several contacts. A list of names is not buying coverage. You need to know who can approve, who can block, who uses the product, and whether each role is evidenced.
Watch out
Do not score the proposal as evidence of customer commitment
Your proposal states what you offered. It does not prove that the buyer accepts the problem, scope, price or decision date.
4. Check the scorecard against the source material
Read the evidence column before you read the recommendation. The most useful output is often a precise gap, such as “economic buyer not identified”, rather than the final label.
Use this check before you update the CRM:
- Open every source cited for a score of
2. - Confirm that it records a customer statement, not an internal interpretation.
- Check the date. A priority from an old discovery call may no longer be current.
- Look for conflicts, such as a target close date in the CRM but no scheduled buyer meeting.
- Assign each next action to one person and give it a date.
- Record the decision and the reason in the opportunity notes.
Check
A usable scorecard has traceable evidence
You should be able to point from every high score to a dated call note, email or customer document. If you cannot, reduce the score or mark it unknown.
5. Make the decision from the gaps
Do not use the total as an automatic verdict. Use it with your non-negotiable criteria. A deal can have a reasonable total while lacking a buyer-led next step, a credible problem, or a route to a decision-maker.
| If the scorecard shows | Decision | What to do next |
|---|---|---|
| Evidence across core criteria, named stakeholders and a dated buyer action | Pursue | Run the agreed next meeting and close the highest remaining gap |
| A credible problem but missing decision, stakeholder or timing evidence | Pause | Set a requalification action. Remove it from committed activity until evidence arrives |
| No confirmed problem, no access path, repeated missed next steps, or a buyer choosing no change | Disqualify | Record the reason, close or downgrade the opportunity, and preserve a re-entry condition if one exists |
For a pause, write the condition that would restart pursuit. For example: “Resume when the operations lead confirms the evaluation process and introduces the budget owner.” This keeps a paused deal from becoming a vague follow-up task.
When the scorecard does not work
If the output is generic, the deal pack is usually too thin or the criteria are too vague. Add the actual call notes and define what each score means. If it invents details, repeat the instruction to use only supplied evidence, then remove any unsupported rows yourself.
If every deal scores highly, tighten the definition of 2 so it requires a dated, customer-sourced statement. If managers disagree with the decision, compare the evidence standard first, not the wording of the recommendation. Update the shared template after the review, then use the revised version on the next active opportunity.
