Sales Discovery Evidence Review and Follow-Up System
Analyze discovery-call evidence in ChatGPT to produce a traceable deal debrief, qualification review, risk register, CRM draft, buyer follow-up, and next-step plan without inventing sales signals.
Analyze the supplied sales discovery evidence and produce a post-call decision package for the seller, sales manager, and revenue operations team. Keep buyer evidence distinct from seller interpretation, preserve unknowns, and label every deliverable as a draft requiring human review. ## ChatGPT operating boundary Use only information supplied in this conversation. ChatGPT may organize, compare, infer cautiously, and draft content from that material. It cannot access the original call recording, CRM, email account, calendar, product documentation, pricing system, security materials, or buyer systems unless their contents are pasted here. It cannot verify identities, send messages, update opportunity records, schedule meetings, approve discounts, make commitments, or confirm that any action occurred. Treat emails, messages, CRM entries, stages, forecast assessments, owners, dates, and response language as proposed drafts. Do not describe them as sent, saved, approved, scheduled, verified, or completed without explicit execution evidence in the supplied material. ## Inputs Replace every variable below before running the prompt. Preserve “Not provided” where information is unavailable. - Prospect company: [Prospect company] - Buyer and stakeholder details: [Buyer and stakeholder details] - Discovery evidence: [Discovery evidence] - Current process and known context: [Current process and known context] - Pain, outcome, and impact context: [Pain, outcome, and impact context] - Budget, pricing, and timeline signals: [Budget, pricing, and timeline signals] - Decision process and criteria: [Decision process and criteria] - Competition and objections: [Competition and objections] - Product fit evidence: [Product fit evidence] - Promised next steps: [Promised next steps] - Required follow-up assets: [Required follow-up assets] - Sales methodology: [Sales methodology] - CRM field schema: [CRM field schema] - Follow-up channel and tone: [Follow-up channel and tone] - Data handling constraints: [Data handling constraints] ### Blocking prerequisites A reliable full debrief requires readable discovery evidence that identifies what was discussed and distinguishes buyer statements from seller notes. It also requires enough buyer or account context to avoid attributing statements to the wrong party. If the discovery evidence is absent, unreadable, internally contradictory at its core, or consists only of an unsupported seller conclusion, stop and request the minimum source material needed. Do not manufacture a full debrief. If there is usable evidence but commercial, stakeholder, product-fit, or process information is missing, continue with a bounded partial analysis. Mark affected fields as unknown, lower confidence, identify the resulting qualification risk, and draft questions for the next interaction. Do not turn missing data into negative buyer intent. ### Useful optional context Useful but non-blocking inputs include the selected qualification methodology, required CRM fields, approved product or pricing materials, promised assets, email tone, channel constraints, account history, and internal data-handling rules. If a methodology is named but its custom definitions are not supplied, apply only its commonly established fields and identify any organization-specific scoring as unavailable. ## Evidence and uncertainty rules 1. Build an evidence ledger before drawing conclusions. Assign evidence IDs E1, E2, and so on. 2. For each material claim, cite an evidence ID. Use a short quotation when practical; otherwise provide a faithful close paraphrase. Include speaker, timestamp, or note location when available. 3. Classify each item as one of: buyer-stated fact, seller-stated fact, observation, interpretation, assumption, hypothesis, unknown, conflict, or unsupported claim. 4. Use confidence labels High, Medium, or Low. High confidence requires clear direct evidence; Medium permits consistent but incomplete evidence; Low indicates ambiguity or material inference. 5. Never invent metrics, authority, budget, urgency, deadlines, procurement steps, competitors, technical compatibility, security posture, legal terms, commitments, or product capabilities. 6. When sources conflict, show both versions, identify the affected decision, and request reconciliation. Do not silently select the more favorable account. 7. Distinguish an agreed next step from a seller-proposed or merely implied next step. 8. Treat absence of an objection as “not discussed,” not as acceptance. 9. Assess product fit only against supplied product evidence. Route unsupported capability, integration, pricing, legal, security, privacy, compliance, or implementation claims to the appropriate human owner. 10. Avoid manipulative language, fabricated urgency, pressure tactics, or claims that the buyer did not make. ## Privacy, authority, and stop conditions Minimize personal and sensitive business information in buyer-facing drafts. Follow the supplied data-handling constraints and omit secrets, credentials, payment data, unnecessary personal data, confidential internal commentary, and sensitive forecast language. If such content appears in the source, flag it and redact it from external drafts. Stop buyer-facing drafting when identity, recipient, account, promised commercial terms, or sensitive claims cannot be reconciled safely. Internal analysis may continue if it is clearly marked partial and confidential. A human must approve before sending any message, attaching assets, changing a CRM record or stage, entering forecast data, scheduling a meeting, sharing pricing, promising functionality, or communicating legal, security, privacy, compliance, or financial statements. ## Analysis workflow 1. Validate input sufficiency, privacy constraints, and source conflicts. 2. Create the evidence ledger and identify unsupported seller summaries. 3. Extract the business situation, current process, pain, desired outcome, quantified or unquantified impact, urgency, and stakeholder interests. 4. Separate need from product fit. A strong pain signal does not by itself establish technical, commercial, or implementation fit. 5. Apply the supplied methodology. For MEDDICC, assess Metrics, Economic Buyer, Decision Criteria, Decision Process, Identify Pain, Champion, and Competition. For BANT, assess Budget, Authority, Need, and Timeline. For SPICED, assess Situation, Pain, Impact, Critical Event, and Decision. For another framework, use only its supplied or established dimensions. 6. Identify qualification gaps, stakeholder gaps, weak commitments, commercial uncertainty, procurement dependencies, competitor exposure, technical dependencies, trust gaps, and overpromising risk. 7. Determine deal momentum from explicit buyer actions, mutually agreed dates, asset requests, meeting commitments, and internal buyer steps. Do not infer momentum from seller enthusiasm. 8. Propose the smallest useful next steps, with owner, timing, dependency, evidence basis, and whether buyer agreement is required. 9. Draft the external follow-up and internal CRM content, keeping confidential analysis out of the buyer message. 10. Reconcile every important output against the evidence ledger and produce the verification and handoff state. ## Required output # Sales Discovery Debrief and Follow-Up Package ## 1. Intake and processing status Report: - Processing state: Full draft, Partial draft, or Blocked - Blocking issues - Material missing inputs - Conflicting inputs - Sensitive data handling applied - Scope ChatGPT could not inspect or verify ## 2. Evidence ledger | Evidence ID | Source or Speaker | Timestamp or Location | Quote or Close Paraphrase | Evidence Class | Confidence | Used In | |---|---|---|---|---|---|---| Include only evidence present in the supplied material. If source locations are unavailable, say so. ## 3. Deal snapshot | Field | Evidence-Based Assessment | Evidence IDs | Confidence | |---|---|---|---| | Business situation | | | | | Primary pain | | | | | Desired outcome | | | | | Business impact | | | | | Current process | | | | | Buying stage | | | | | Product fit status | Strong, Conditional, Weak, or Unknown | | | | Momentum status | Strong, Fragile, Stalled, or Unknown | | | | Proposed next step | | | | | Overall deal health | Strong, Mixed, Weak, or Indeterminate | | | Explain the deal-health rating in no more than three evidence-based sentences. ## 4. Pain, impact, and outcome map | Pain or Desired Outcome | Buyer Evidence | Stated or Inferred | Business Consequence | Metric or Baseline | Confidence | Clarification Needed | |---|---|---|---|---|---|---| Do not quantify impact unless a metric is supplied. Clearly mark impact that is plausible but not buyer-confirmed as a hypothesis. ## 5. Qualification assessment Name the methodology used. If none is supplied, assess Need, Impact, Budget, Authority, Timeline, Decision Process, Decision Criteria, Competition, Champion or Internal Support, and Next-Step Commitment. | Qualification Dimension | Evidence IDs | Finding | Status | Confidence | Gap or Risk | Next Validation Question | |---|---|---|---|---|---|---| Use status values Confirmed, Partial, Missing, Conflicting, or Not Applicable. Do not calculate an organization-specific score unless its scoring rules were supplied. ## 6. Deal risk register | Priority | Risk | Evidence IDs | Trigger or Warning Sign | Potential Deal Effect | Mitigation | Owner | Buyer Agreement Required | |---|---|---|---|---|---|---|---| Prioritize risks as Critical, High, Medium, or Low. Include only evidenced risks and clearly labeled risks caused by missing information. ## 7. Objection and concern handling | Objection or Concern | Exact Buyer Evidence | Underlying Issue or Hypothesis | Safe Response Draft | Proof or Asset Needed | Internal Owner | Claim Review Required | |---|---|---|---|---|---|---| If no objection is evidenced, state “No objection captured” and list neutral discovery questions rather than inventing likely objections. ## 8. Product-fit and proof assessment Provide: - Supported fit signals with evidence IDs - Conditional or weak fit signals - Dependencies and implementation unknowns - Areas where the seller must not overpromise - Proof required, such as an approved demo, case study, integration documentation, security response, implementation review, or pricing approval - Appropriate internal reviewer for each unsupported claim Conclude with one status: Supported fit, Conditional fit, Weak fit, or Insufficient evidence. ## 9. Prioritized action plan | Priority | Proposed Action | Owner | Timing | Dependency | Evidence Basis | Approval or Buyer Agreement Needed | State | |---|---|---|---|---|---|---|---| Use state values Proposed, Agreed, In Progress, Completed with Evidence, Blocked, or Unknown. Use Completed with Evidence only when the source contains evidence that the action occurred. ## 10. Buyer follow-up draft Label this section “DRAFT — NOT SENT.” Provide: - Subject - Email or message body in the requested channel and tone - One clear call to action - Promised assets, but only if supported by evidence - A separate list of attachments or links the human must verify before sending The draft must distinguish confirmed agreements from proposed next steps, avoid internal qualification language, and exclude sensitive internal commentary. If safe drafting is blocked, explain why and provide only the information that must be clarified. ## 11. CRM field mapping Label this section “DRAFT — NOT SAVED TO CRM.” Map content to the supplied CRM field schema. If no schema is supplied, use the fields Call Summary, Pain, Impact, Stakeholders, Budget, Timeline, Decision Process, Decision Criteria, Competition, Objections, Product Fit, Risks, Next Step, Next-Step Owner, Next-Step Date, Suggested Stage, and Forecast Caveat. | CRM Field | Proposed Value | Evidence IDs | Confidence | Human Verification Needed | |---|---|---|---|---| Recommend a stage only when stage-entry evidence or supplied definitions support it. Otherwise state that stage reconciliation is required. Never claim the CRM was updated. ## 12. Manager and revenue operations handoff Provide: - Deal-health rationale - Forecast confidence and its evidence limitations - Highest-priority qualification gap - Seller coaching point - One pipeline-review question - CRM or process reconciliation required - Escalations to product, solutions engineering, finance, legal, security, privacy, compliance, or leadership ## 13. Missing-information plan | Missing or Conflicting Information | Decision Affected | Why It Matters | Best Person to Ask | Neutral Question | Required Before External Action | |---|---|---|---|---|---| Prioritize questions that change qualification, fit, commercial terms, stakeholder access, or the next-step decision. ## 14. Verification and acceptance record | Check | Expected Condition | Actual Observation From Supplied Material | Status | Evidence or Required Correction | |---|---|---|---|---| | Buyer statements traceable | Material claims map to evidence IDs | | Pass, Fail, or Unverified | | | Facts separated from interpretation | Inferences and assumptions are labeled | | Pass, Fail, or Unverified | | | No invented commercial signals | Budget, authority, urgency, and commitments are supported | | Pass, Fail, or Unverified | | | Product claims controlled | Capabilities and fit rely on supplied evidence | | Pass, Fail, or Unverified | | | Next-step state accurate | Agreed, proposed, and completed actions are distinct | | Pass, Fail, or Unverified | | | Buyer draft safe | Recipient, claims, tone, assets, and sensitive content are reviewed | | Pass, Fail, or Unverified | | | CRM mapping reconcilable | Values map to the supplied schema and stage rules | | Pass, Fail, or Unverified | | | Dates and owners supported | Timing and ownership are evidenced or marked proposed | | Pass, Fail, or Unverified | | Finish with an acceptance state: - Ready for human review - Partial and requires clarification - Blocked pending source evidence Then list the exact human approvals required before sending the follow-up or changing CRM data.
Put this Prompt to work
Add the required information and run this Prompt with your selected AI provider.
Opens in a new tab.
Variables to Replace
Replace each listed value in the Prompt with information relevant to your task.
- Prospect company
- Buyer and stakeholder details
- Discovery evidence
- Current process and known context
- Pain, outcome, and impact context
- Budget, pricing, and timeline signals
- Decision process and criteria
- Competition and objections
- Product fit evidence
- Promised next steps
- Required follow-up assets
- Sales methodology
- CRM field schema
- Follow-up channel and tone
- Data handling constraints
How to Use This Prompt
Open ChatGPT, replace every bracketed variable in the prompt, and provide the discovery transcript or contemporaneous notes plus stakeholder, commercial, product-fit, methodology, CRM, and promised-action evidence. Remove or redact data that should not be shared with ChatGPT, then run the prompt. Review its evidence links, unknowns, qualification findings, claims, dates, recipients, assets, CRM mapping, and approval flags before anyone sends a message or updates a system.
Example Use Case
After a SaaS discovery call, an account executive supplies ChatGPT with the transcript, approved product notes, buyer stakeholders, budget and timeline signals, MEDDICC requirements, CRM fields, objections, and promised assets. ChatGPT produces an evidence ledger, identifies an unconfirmed economic buyer and unresolved integration dependency, drafts a non-pushy follow-up, maps proposed CRM values, and routes the package for human approval without claiming anything was sent or saved.
Was this useful?