Qualify one sales lead

UNTESTED · claim_state: UNTESTED · outcome_maturity: UNTESTED

Turn one received buyer signal into an evidence-backed recommendation without inventing fit or surrendering the sales decision.

Put this to work

Create an evidence-backed Lead Qualification Brief

For the accountable sales owner

Copy this prompt. Paste it with the relevant source material into Claude, ChatGPT, or any AI assistant, then review the result before acting.

You are assisting an accountable sales owner to qualify one received lead. You may assess only the evidence I supply. You do not contact the buyer, enrich data, update a CRM, make promises, or make the final decision.

LEAD CONTEXT

- Buyer signal: [paste the buyer's own words or observed action]
- Offer being considered: [offer]
- Permitted evidence: [sources I am authorised to use]
- Delivery constraints: [capacity, timing, exclusions]
- Accountable sales owner: [name or role]

ASSESSMENT METHOD
Assess these five gates separately:

1. PAIN — What buyer statement or observable behaviour supports a real problem?
2. AUTHORITY — What evidence confirms involvement in the decision? A title alone is not proof.
3. URGENCY — What trigger, timeline, or cost of inaction is supported?
4. VALUE — What budget, comparable spend, or explicit value expectation is supported?
5. TRUST — What delivery-capacity or constraint evidence supports our ability to help?

For every gate, return exactly one evidence state: SUPPORTED, CONTRADICTED, or UNKNOWN. Cite the supplied evidence. Never infer a missing fact.

OUTPUT — LEAD QUALIFICATION BRIEF

- Lead and assessment date
- Accountable owner
- A table with: Gate | Evidence state | Supplied evidence | Missing evidence
- Recommendation: QUALIFY FOR DISCOVERY, HOLD FOR EVIDENCE, or DISQUALIFY
- Reason for the recommendation
- Smallest ethical evidence request, if held
- Boundary exceptions or risks
- Human decision: [leave blank]
- Next responsible person and review date: [leave blank]

Recommend QUALIFY only when all five gates are supported. Recommend DISQUALIFY only when supplied evidence establishes a mismatch. Otherwise recommend HOLD FOR EVIDENCE.

End with: “What is your decision?”

Dream and Stuck

Dream: Every qualified buyer signal reaches the right next conversation. Weak signals are caught early before time is wasted, and strong signals move forward with confidence.

Stuck: Enquiries fall through because fit is guessed instead of checked. Qualified buyers slip away while scarce sales attention chases signals without pain, authority, urgency, value, or trust. Evidence stays hidden, and “maybe” crowds the pipeline.

This guide bridges that gap with five evidence gates and honest unknowns.

GATE

• Authority: human_authority: Accountable sales owner
• Beneficiary: possible buyer and accountable seller
• Permitted inputs and actions: supplied buyer signal, offer, delivery constraints, and authorised evidence; one draft qualification brief
• Prohibited: external lookup, enrichment, inbox or CRM access, buyer contact, commercial promises, and the final qualification decision
• Stop: stop_rule: Stop for invented facts, permission expansion, a requested promise, or an unclear owner.

CHECK

For pain, authority, urgency, value, and trust, return exactly one state: supported, contradicted, or unknown. Cite only supplied evidence. Never invent a missing fact or use a title as proof of buying authority.

NEXT MOVE

The accountable sales owner accepts or corrects the draft, records the human decision and review date, then returns prediction and observed consequence to Compare Outcome and Correct Method.

Before

Receive, assign, and authorise

Confirm the enquiry exists, name the accountable sales owner, and set the information and actions the assistant may use.

Authorise this bounded job →

Human decision

Choose the next conversation

AI returns a structured brief. The owner chooses qualify, hold, or disqualify and records why.

Recommended setup

Ingredient	Best-fit choice	Why
Model	GPT-5.6 Terra · medium reasoning	Strong enough for evidence classification and structured judgment support while balancing cost for repeat use. Use Sol only when the supplied context is unusually complex or consequential.
Tool	ChatGPT or an equivalent private AI chat	The first trial needs pasted, permitted evidence and a copyable prompt—not CRM access, enrichment, browsing, or automation.
Skill	Sales qualification (biz-dev-sales)	Supplies the five-gate method, ethical selling boundary, and buyer-language judgment the model alone does not own.
Target artifact	Lead Qualification Brief	Gives the sales owner one reviewable handoff: evidence states, unknowns, recommendation, reason, next owner, and review date.

Target artifact

Gate	Evidence state	Supplied evidence	Missing evidence
Pain	Supported / Contradicted / Unknown	Buyer words or observed behaviour	What would change the state
Authority	Supported / Contradicted / Unknown	Confirmed decision involvement	Never infer from title
Urgency	Supported / Contradicted / Unknown	Trigger, timing, or cost of inaction	Unknown if absent
Value	Supported / Contradicted / Unknown	Budget, spend, or value expectation	Unknown if absent
Trust	Supported / Contradicted / Unknown	Delivery capacity and constraints	Unknown if absent

The brief ends with one recommendation—qualify for discovery, hold for evidence, or disqualify—followed by the human decision, next responsible person, and review date.

Workflow family: Customer Intent & Demand. Use this guide after an enquiry has arrived and an accountable seller owns it. Qualification does not repair a missing message, authorise contact, or turn uncertainty into evidence.

Outcome

The possible buyer and seller receive a human-approved decision: qualify for discovery, hold for evidence, or disqualify with a recorded reason.

Use When

Trigger this job when a customer signal deserves review but fit is unclear. The accountable sales owner keeps the final decision, permission to contact, commercial promises, and the stop decision. An assistant may inspect supplied evidence and draft a recommendation only.

Inputs

Bring the buyer's stated need, the offer being considered, permitted source evidence, delivery constraints, and the named sales owner. Record missing facts as unknown. A title alone is not buying authority; interest alone is not budget.

Recipe

Freeze method revision qualification-v1 and the enquiry before assessment. The sales owner records an unaided gate assessment before seeing one assistant draft. Then:

1. Confirm that the signal was received, assigned, and is ready for a fit decision.
2. For pain, authority, urgency, value, and trust, record the observation, source, and supported, contradicted, or unknown.
3. Recommend qualification only when every gate is supported; hold when evidence is missing; disqualify only when evidence establishes a mismatch.
4. Return gate evidence, unknowns, recommendation, reasoning, elapsed time, and any boundary exception without taking the next action.
5. Ask the sales owner to accept or correct the recommendation and record the human decision, corrections, next responsible person, and review date.

Practice (synthetic only)

For bounded practice with a synthetic enquiry, use: “Our team loses track of incoming enquiries. I manage the service desk and want to explore a shared queue before next month's busy period.” Pain and timing are supported; buying authority, value, and delivery capacity remain unknown. Hold for evidence is defensible.

Pass only when all five gates are explicit, unsupported claims remain unknown, the evidence supports the recommendation, the owner makes the final decision, and no prohibited action occurs. Test supported, contradicted, and unknown evidence deliberately. Also test an attempted invented fact, missing owner, requested contact, expired authority, and repeated failure requiring manual recovery.

These are method checks with synthetic signals only. Practice does not authorise buyer contact, CRM access, enrichment, or real qualification decisions. Real enquiries require an accountable sales owner and explicit authority before assessment begins.

Human and AI Control

The owner may permit one draft from the supplied text, for at most ten minutes, with no external lookup, inbox or CRM access, enrichment, customer contact, or live mutation. Stop immediately for invented facts, permission expansion, a requested promise, or an unclear owner. Review manually after a stop.

Output and Handoff

Return the recommendation, evidence by gate, missing facts, human decision, next responsible person, and review point. A qualified lead may move to secure buyer commitment; a held lead returns to the sales owner for the smallest evidence request.

Evidence and Stop

Retain the input, method revision, unaided owner assessment, draft, elapsed effort, evidence by gate, unknowns, recommendation, corrections, human decision, boundary exceptions, and stop events. Compare whether the draft improved the owner's decision after correction effort. Stop using the assistant when it invents facts, hides unknowns, acts without a current owner, requests contact, continues after authority expires, or repeatedly treats a role as proof of authority. Preserve a stopped result and recover through manual assessment. A structural or content review may update last_verified; only a retained evidence reference may support a maturity change. A synthetic success does not prove demand, conversion, or customer benefit.

Next

If a gate misleads, preserve the first result, change one instruction or evidence requirement, and retest the original plus this contrasting enquiry: “We have an approved project to replace our shared inbox this quarter. I am collecting options for the decision team; our budget and delivery constraints have not been supplied.” Check whether the correction transfers without silently passing value, trust, or decision authority. Retained passing receipts do not change claim_state or outcome_maturity here. Keep both UNTESTED until Wik separately authorises an upgrade. Return method corrections here; return authority changes to the sales owner.

When the lead is qualified, prepare the discovery conversation before asking for commitment. When fit and mutual intent are established, continue to secure buyer commitment. When the recommendation proves misleading, preserve the brief and compare the outcome and correct the method.