---
name: prospect-intelligence-deep-dive
description: Builds a deep profile on a prospect covering background, role and priorities, company context, and buying signals, ending in one recommended outreach angle. Use when the user asks to "do a deep dive on this prospect", "research this person before I reach out", "full profile on [name]", "what's the angle for this prospect", or wants outreach-ready research on a person.
license: MIT
compatibility: Requires web search; connected CRM optional for relationship context. Read-only.
metadata:
  author: Builder.org
  version: 2.0.0
  persona: Strategic Steve, Marketing Meg
  rave: R4 A3 V3 E3
  category: crm
---

# Prospect Intelligence Deep-Dive

## What it does

A deeper-than-a-brief profile of one person you want to reach: who they are, what they are measured on, what is happening at their company, and the sharpest way to open. The deliverable is an angle you could paste into the first line of an email, backed by a dossier that justifies it.

## When to use

- "Deep dive on [prospect]", "research this person before I reach out", "full profile on [name]".
- "What's the angle for this prospect", "why would [contact] take a meeting right now".
- Not for someone you already have history with before a booked meeting; that is `pre-meeting-brief`. If you have a LinkedIn URL and want their own words mined at full depth, `prospect-research` goes further on quotes.

## Inputs

A prospect: name plus company, or an email, or a profile link. Optional: the offer you are bringing, which turns a generic angle into a fitted one.

## The bar

- Verified facts and inferences never blend. A fact has a source; an inference says "likely" and names what it rests on.
- The recommended angle ties to one dated, specific signal. "They're growing" is not a signal; "they posted six SDR openings in January" is.
- Priorities are stated as what this person is measured on, not what their title suggests they might enjoy.
- Thin footprint means saying confidence is low, not writing fiction with hedges.

## Workflow

1. **Profile the person.** Role, seniority, tenure, career path, and public activity (posts, talks, interviews). From these, infer the two or three numbers this person is probably accountable for, and say why you think so.
2. **Profile the company.** Size, model, funding stage, recent news, hiring pattern, visible tech choices. Favor the last 90 days; date everything.
3. **Pull relationship context** from your CRM when connected: prior touches, colleagues already in the CRM, any dead deal with their company and why it died. A past loss is often the best opener there is.
4. **Detect buying signals.** New role under 90 days, fresh funding, team expansion, a public complaint about a problem you solve, a competitor stumble. Rank signals by recency times relevance; keep the top three.
5. **Commit to the angle.** One opener tied to the best signal, written as the actual first line you would send, plus two or three supporting talking points. If an offer was given, connect the signal to the offer in one step, not three.
6. **Assemble the dossier** in the output format.

## Output

```
## [Name] prospect intelligence
**Role:** [title · seniority · tenure] · **Likely measured on:** [2-3 inferred KPIs]
**Company:** [size · industry · stage]

### Buying signals
- [Dated signal with source]

### Relationship context
[Prior touches, mutual colleagues, past deals, or "none in the CRM"]

### The angle
Opener: "[the first line you would actually send]"
Why it works: [one sentence tying it to the signal]

### Talking points
1. [Point tied to a signal or priority]
2. [Point tied to a signal or priority]

### Confidence
[What is verified, what is inferred, what is unknown]
```

## Worked example

```
Fictional example. Never reuse these names or numbers in real output.

## Dana Okafor prospect intelligence
**Role:** VP of Operations at Brightpath Staffing · senior · 4 months in role · **Likely measured on:** time-to-fill, recruiter utilization, margin per placement
**Company:** ~85 employees · healthcare staffing · bootstrapped, 11 years old

### Buying signals
- Jan 12: promoted from Director to VP after the prior VP left (LinkedIn)
- Jan 27: Brightpath posted 5 recruiter openings across two offices (careers page)
- Feb 8: Dana commented on a staffing-industry post that "our credentialing turnaround is the bottleneck nobody talks about" (LinkedIn)

### Relationship context
None in the CRM. A colleague, Marcus Lee (recruiting lead), downloaded a pricing guide in November and never replied to the follow-up.

### The angle
Opener: "You said credentialing turnaround is the bottleneck nobody talks about. We built our whole onboarding flow around that sentence."
Why it works: it quotes her own public words back within a month of her writing them, in her first quarter in a role where she has to show a number moving.

### Talking points
1. New VPs get about two quarters to pick their first visible win; credentialing time is measurable fast.
2. Five open recruiter seats means throughput pressure lands on ops, not just recruiting.
3. Marcus already looked at pricing in November, so there is internal awareness to build on, gently.

### Confidence
Verified: role change, job postings, her comment (all dated and public). Inferred: her KPIs, the two-quarter window. Unknown: budget authority; her title suggests it but the org chart is not public.
```

## Never

- Never invent a quote, a personal detail, or a mutual connection. One fabricated line poisons the whole dossier.
- Never present an inference as a fact. "Likely" is not a weakness; it is the honesty that makes the facts credible.
- Never deliver three angles. The dossier exists to make one decision.
- Never use personal-life details (family, health, politics) as an angle, even when public.

## Recovery

- Thin personal footprint: shift weight to company-level signals, say so, and make the angle role-based ("VPs of Ops in staffing usually inherit X") rather than person-based.
- No company news either: the angle becomes a sharp problem statement for their segment, honestly framed as a hypothesis to test in the first reply.
- Cannot disambiguate two people with the same name: ask one question, or if you cannot ask, pick the one matching the company given and state that assumption first.
- Signal contradicts the offer (they just bought a competitor): say so plainly and recommend the nurture path instead of forcing an opener. An honest "not now" is a good output.