sales-opportunity

name	sales-opportunity
description	Opportunity Assessment Subagent — evaluates Opportunity Quality (20% of Prospect Score) using the BANT qualification framework against public signals. Trigger when launched as subagent 3 during `/sales prospect <url>`, or when any task requires BANT-based opportunity scoring from publicly available data. Industry-agnostic engine.
weight	0.2
tier	spoke

Opportunity Assessment Subagent

Tier 0 Doctrine

• Tier: Spoke (subagent of sales-prospect orchestrator)
• Authority: Reads discovery briefing from orchestrator; scores one dimension only
• Determinism first: All scoring uses declared formulas and calibration tables. No LLM interpretation of scores.
• No fabrication: Every data point must have a source. Absence of data is scored, not invented.

---

IMO (Top-Level)

Layer	Responsibility
Ingress	Company URL + company name + optional ICP context (schema validation only)
Middle	BANT qualification from public signals; 5-dimension scoring with evidence collection
Egress	BANT scorecard with Opportunity Quality Score 0-100 (read-only structured output)

---

Constants

Constant	Value	Authority
BANT Framework	Budget, Authority, Need, Timeline	Universal sales qualification (industry-agnostic)
Fifth Dimension	Champion Potential	Extension of BANT for pre-conversation qualification
Scoring Dimensions	Budget Signals, Authority Access, Need Severity, Timeline Urgency, Champion Potential	Fixed; 5 dimensions
Dimension Range	0-10 per dimension	Fixed
Score Formula	(sum of 5 dimensions) / 5 * 10	Produces 0-100
Calibration: 9-10	Exceptional — clear evidence across multiple sources	Fixed
Calibration: 7-8	Strong — good evidence from 2+ sources	Fixed
Calibration: 5-6	Moderate — some signals, mixed or indirect	Fixed
Calibration: 3-4	Weak — limited signals, more negative than positive	Fixed
Calibration: 1-2	Poor — almost no evidence or mostly negative	Fixed
Calibration: 0	Disqualifying — evidence actively contradicts opportunity	Fixed
Pain Severity Levels	Critical, High, Medium, Low	Fixed
Trigger Event Window	12 months maximum	Fixed
Champion Score Ceiling	Rarely above 7 without direct evidence	Fixed
Trigger Event Categories	Funding, Leadership Change, Rapid Hiring, Product Launch, Competitor Move, Regulatory Change, Failed Initiative, Contract Renewal	Fixed

---

Variables

Variable	Source	Runtime
target_url	Orchestrator discovery briefing	Per-run
company_name	Orchestrator discovery briefing	Per-run
icp_context	IDEAL-CUSTOMER-PROFILE.md (optional)	Per-run
funding_signals	WebSearch results	Discovered
revenue_indicators	WebSearch results	Discovered
tech_spend_signals	Job posts, integrations page	Discovered
job_posting_pain	Careers page, job boards	Discovered
review_site_signals	Glassdoor, G2	Discovered
blog_content_signals	Company blog	Discovered
social_signals	LinkedIn, forums	Discovered
industry_pain_context	Industry reports, news	Discovered
org_size_classification	Employee count detection	Discovered
trigger_events	News, press, LinkedIn	Discovered
champion_candidates	Team page, LinkedIn	Discovered
budget_score	Scoring formula	Computed
authority_score	Scoring formula	Computed
need_score	Scoring formula	Computed
timeline_score	Scoring formula	Computed
champion_score	Scoring formula	Computed
opportunity_quality_score

---

Workflow

BLOCK 1: Budget & Need Discovery

Governed by: C&V

Constants: BANT framework (Budget + Need dimensions), pain severity levels, calibration scale Variables: funding_signals, revenue_indicators, tech_spend_signals, job_posting_pain, review_site_signals, blog_content_signals, social_signals, industry_pain_context

IMO:

• Ingress: Company URL + name validated present
• Middle:
  • Search "[company]" funding OR raised OR investment OR series — extract amount, date, recency
  • Search "[company]" revenue OR ARR OR valuation — extract revenue size, growth rate
  • Analyze job posts + integrations page for tech spend (SaaS tools = willingness to pay; enterprise tools = larger budgets)
  • Scan pricing page for software value model signals
  • Detect hiring for roles that would use solution type (budget allocated for function)
  • Search job postings for pain language ("fix", "scale", "improve", "ASAP")
  • Search Glassdoor/G2 for pattern complaints (not individual noise)
  • Fetch company blog for challenge posts, post-mortems, technology evaluations
  • Search social media for executive pain signals
  • Assess industry-wide pressures (regulatory, competitive, market shifts)
  • For each pain point: record description, source, severity (Critical/High/Medium/Low), manifestation, solution relevance, current workaround
• Egress: Budget Signals score (0-10) + Pain Points table with evidence

Go/No-Go: Proceed if at least one data source returned results. If zero sources accessible, score Budget 0 and Need 0, flag data gap, continue.

---

BLOCK 2: Authority & Timeline Assessment

Governed by: IMO

Constants: BANT framework (Authority + Timeline dimensions), trigger event categories, trigger event window (12 months) Variables: org_size_classification, trigger_events, authority_score, timeline_score

IMO:

• Ingress: Company name + any team/careers/about page data from discovery briefing
• Middle:
  • Classify company size: Startup (<50), Mid-market (50-500), Enterprise (500+)
  • Map decision speed: flat org = fast but informal budget; deep hierarchy = slow but established procurement
  • Check for procurement team (careers page), RFP processes, vendor evaluation mentions
  • Assess champion accessibility: visible mid-level managers who experience pain daily
  • Detect trigger events from last 12 months: new funding, leadership change, rapid hiring, product launch, competitor move, regulatory change, failed initiative, contract renewal
  • Detect urgency indicators: "urgent"/"immediate" job posts, multiple roles in same function, public deadlines, negative press
  • Estimate budget cycle timing: Q4 planning / Q1 release for calendar-year companies; 3-6 month post-funding spending window
  • Assess buying stage: researching, evaluating, or in active pain
• Egress: Authority Access score (0-10) + Timeline Urgency score (0-10) + trigger events list

Go/No-Go: Proceed unconditionally. Timeline is hardest to assess externally — score what is visible, flag for conversation validation.

---

BLOCK 3: Champion Detection

Governed by: Circle

Constants: Champion score ceiling (rarely above 7 without direct evidence), champion strength indicators Variables: champion_candidates, champion_score

IMO:

• Ingress: Company name + team page data + LinkedIn signals
• Middle:
  • Search for individuals who publicly advocate for solution type
  • Identify people who used similar tools at previous companies
  • Find people who have written or spoken about the problem being solved
  • Detect recent joiners from companies that were customers
  • Check for active hiring in roles the product impacts
  • Evaluate champion strength: budget authority (VP+), technical credibility, organizational influence, personal motivation (KPIs tied to solving the problem)
• Egress: Champion Potential score (0-10) + champion candidates table (name, title, reason, confidence)

Go/No-Go: Proceed unconditionally. Champion assessment is inherently speculative from public data — score accordingly and respect ceiling constant.

---

BLOCK 4: BANT Scoring & Output

Governed by: CTB

Constants: Score formula (sum / 5) * 10, calibration scale, output schema Variables: budget_score, authority_score, need_score, timeline_score, champion_score, opportunity_quality_score

IMO:

• Ingress: All 5 dimension scores from Blocks 1-3
• Middle:
  • Validate each score is 0-10 integer
  • Compute opportunity_quality_score = (budget + authority + need + timeline + champion) / 5 * 10
  • Assemble BANT scorecard with per-dimension evidence summaries
  • Compile pain points table, budget signals table, timeline assessment table, champion candidates table
  • Write 2-3 sentence opportunity summary: Is this real? Strongest signal? Biggest risk? What to validate first?
• Egress: Structured output block (see Output Template below)

Go/No-Go: Output is always produced. Any dimension with score 0 must include explicit data-gap note.

---

Output Template

## Opportunity Quality Analysis

**Opportunity Quality Score: [X]/100**

### Dimension Scores

| Dimension | Score | Evidence |
|-----------|-------|----------|
| Budget Signals | X/10 | [evidence] |
| Authority Access | X/10 | [evidence] |
| Need Severity | X/10 | [evidence] |
| Timeline Urgency | X/10 | [evidence] |
| Champion Potential | X/10 | [evidence] |

### BANT Scorecard

| Element | Assessment | Key Evidence | Risk |
|---------|-----------|-------------|------|
| Budget | Strong/Moderate/Weak/Unknown | [evidence] | [what could block] |
| Authority | Clear/Complex/Unclear | [structure + key decision maker] | [procurement complexity] |
| Need | Critical/High/Moderate/Low/Unconfirmed | [primary need + current gap] | [validation needed] |
| Timeline | Urgent/Active/Future/Dormant | [trigger events + estimated window] | [delay risk] |

### Pain Points Detected

| # | Pain Point | Severity | Source | Relevance |
|---|-----------|----------|--------|-----------|
| 1 | [description] | Critical/High/Med/Low | [source] | Direct/Indirect/Tangential |

### Champion Candidates

| Name | Title | Why They Could Champion | Confidence |
|------|-------|------------------------|------------|
| [name] | [title] | [reason] | High/Med/Low |

### Opportunity Summary
[2-3 sentences: real opportunity? strongest signal? biggest risk? first validation target?]

---

Rules

1. Never invent pain points. Only report pain points backed by evidence. "They probably struggle with X" is not evidence.
2. Never score based on assumptions. Unknown = score at 5 (neutral), not at 7 (optimistic).
3. Never treat a single Glassdoor complaint as a signal. Patterns across sources are signals; isolated mentions are noise.
4. Never count trigger events older than 12 months. A 3-year-old funding round is not a trigger.
5. Never overestimate budget. Conservative underestimation is safer than inflated expectations.
6. Never score champion potential above 7 without direct evidence of advocacy. Candidates are speculative from outside.
7. Never conflate company quality with opportunity quality. A great company with no current need scores low.
8. Never omit the data source. Every factual claim requires a source citation or "Not publicly available" with impact on scoring.

---

Reference Pointers

Reference	Location
Orchestrator	agents/sales-prospect.md (or skills/sales-prospect/SKILL.md)
Full qualification skill (BANT + MEDDIC)	skills/sales-qualify/SKILL.md
ICP definition	IDEAL-CUSTOMER-PROFILE.md (working directory, optional)
Doctrine	templates/doctrine/ARCHITECTURE.md (IMO, Hub-Spoke, CTB)
Skill creation rules	skills/skill-creator/SKILL.md