# Company Deep Dive

> Multi-wave, multi-agent deep company research methodology. 4 parallel research lanes, 3 waves, 15-45 minutes to a full intelligence brief. Be the most prepared person in the room.

- Skill: `puretechnyc/company-deep-dive` (Agent Skill)
- Install (CLI): `npx skillmds@latest add puretechnyc/company-deep-dive`
- Raw SKILL.md: https://api.skillmd.com/api/skills/puretechnyc/company-deep-dive/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: puretechnyc (https://skillmd.com/u/puretechnyc)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/puretechnyc/company-deep-dive

---


# Company Deep Dive

## What This Builds

A research brief that makes you the most prepared person in the room before a sales pitch, partnership discussion, or business engagement.

Not a summary of their website. Not industry boilerplate. Real intelligence:

- **Real company size/revenue/locations** -- triangulated from permits, employee count, review velocity
- **Tech stack assessment** -- what software they are actually running (or not)
- **Top pain points with $ estimates** -- specific to their scale, market, and review patterns
- **Decision-maker profiles** -- names, titles, tenure, how they think
- **THE HOOK** -- their own words about their pain, pulled from job postings, website copy, or review responses
- **Reputation data** -- review counts, ratings, complaint patterns, response behavior

When you walk in with this brief, you are not selling. You are diagnosing.

---

## When to Use

- Before any enterprise sales call or pitch where close rate matters
- Before a first meeting with a potential client
- Before a partnership discussion
- Before competitive analysis
- Any time "I Googled them" is not enough

Time investment: 15-45 minutes depending on depth and company complexity.

---

## The 4-Agent Parallel Research Model

Run these four research lanes in parallel. Each owns a distinct lane. No duplication, no gaps.

### Agent A: General Deep Dive

**Mission**: Paint the full picture -- who they are, what people say about them, and most importantly, what they have said about their own pain.

**Sources to hit**:
- Company website (all pages, especially About, Team, Careers, blog posts)
- Google Maps listing (review count, rating, response patterns, Q&A)
- Review platforms (Yelp, G2, Capterra, BBB, industry-specific)
- LinkedIn company page (employee count trend, recent posts, job activity)
- Social media (community engagement, complaint threads)
- Job postings on Indeed/LinkedIn (THIS IS GOLD -- see The Holy Grail section)
- Recent news mentions, press releases
- Podcast appearances or interviews by leadership

**What to produce**:
- Company narrative (founding story, growth arc, current positioning)
- Review pattern analysis (what customers love, what they complain about)
- Social proof signals (awards, certifications, association memberships)
- THE HOOK (their own words -- mandatory output)

---

### Agent B: Permit and Corporate Research

**Mission**: Get the hard facts -- legal entity, officers, founding date, license status.

**Sources to hit**:
- State contractor license lookup (varies by state and industry)
- Secretary of State / corporate registry
- Permit aggregators (e.g., BuildZoom for construction/trades)
- Government contract databases (USASpending.gov, SAM.gov)

**What to produce**:
- Legal entity name and structure
- Officer names and titles (these are your actual decision-makers)
- Founding date
- License number and status
- Total permit/project volume (if applicable)
- Geographic footprint
- Any disciplinary actions or complaints on record

---

### Agent C: Competitive Landscape

**Mission**: Understand where this company sits in their market.

**Sources to hit**:
- Search "best [service] in [city]" -- who shows up alongside them?
- Review comparison across competitors
- LinkedIn search for similar companies
- Industry ranking sites and association membership lists

**What to produce**:
- Top 3-5 named competitors with basic profiles
- Review comparison table (count, rating, recency)
- Differentiators: where does the target company win? Where do they lose?
- Market position: dominant player, challenger, or niche operator?
- Gaps the competitor set reveals

---

### Agent D: Financial and Scale Signals

**Mission**: Triangulate revenue and true scale from proxy signals.

**Sources to hit**:
- LinkedIn employee count (and trend over time)
- Job postings (count of open roles as growth proxy)
- Public financial databases (Dun & Bradstreet, ZoomInfo)
- Government loan databases
- Physical footprint signals (locations, fleet, real estate)

**What to produce**:
- Revenue triangulation (show your work)
- Employee count estimate
- Growth signal (hiring, new locations = expanding; layoffs, review decline = contracting)

---

## Wave Structure

### Wave 1: Corporate Fundamentals (Agent B only, 5-10 min)

Run Agent B first. Corporate and permit data grounds everything else. You need the legal entity name to search correctly on other platforms.

### Wave 2: Market Position (Agents A + C in parallel, 15-25 min)

Launch A and C simultaneously. The qualitative picture and competitive picture are independent research lanes.

### Wave 3: Financial Triangulation (Agent D, 10-15 min)

Agent D runs last because revenue triangulation requires inputs from the other agents.

---

## The Holy Grail

The single most valuable thing you can find is something the company said about their own pain.

**Why it matters**: When you walk in and say "I noticed you wrote on your website that you can't keep up with demand" -- you are not selling anymore. You are reflecting their reality back to them.

**Where to find it**:

1. **Job postings** -- Every job posting is a company admitting a gap.
   - "Dispatcher needed immediately" = scheduling is breaking
   - "Experienced CSR to handle high call volume" = they are losing calls
   - "Office manager to implement processes" = they have no processes
   - Read the job description body, not just the title.

2. **About page / Our Story copy** -- Look for phrases like:
   - "Demand for our services has grown faster than expected"
   - "We have expanded to [X] locations"
   - These are pride statements that also reveal operational strain.

3. **Owner responses to negative reviews** -- Promises to improve reveal what they are currently failing at.

4. **Their own marketing copy** -- "We answer the phone when others don't" means phone coverage is a known industry problem.

---

## Revenue Triangulation Formula

### Method 1: Volume-Based

```
(Total projects/permits in trailing 12 months) x (average project value)
= Revenue floor estimate
```

### Method 2: Employee Count

```
(Total employee count) x (industry average revenue per employee)
= Revenue midpoint estimate
```

### Method 3: Review Velocity

```
(Total reviews) / (years in business) = annual review rate
(Annual review rate) / (industry review rate %) = implied annual job volume
(Annual job volume) x (average ticket) = revenue estimate
```

### Reporting the Range

Always show your work. Never present a single number.

```
Method 1: $X - $Y
Method 2: $X - $Y
Method 3: $X - $Y
Triangulated range: $X - $Y (confidence level)
```

---

## Research Brief Output Format

```markdown
# [Company Name] -- Research Brief
**Prepared**: [Date]
**Confidence level**: [High / Medium / Low]

## Company Overview
- Legal entity, officers, founded, locations, primary services

## Real Scale
- Employee count, geographic reach, project volume

## Revenue Range
- Triangulated estimate with methods shown

## Tech Stack Assessment
- Detected / Not detected / Suspected (with evidence)

## Top 3 Pain Points
1. [Pain] -- Evidence: [source]. Estimated impact: $[X]
2. [Pain] -- Evidence: [source]. Estimated impact: $[X]
3. [Pain] -- Evidence: [source]. Estimated impact: $[X]

## Decision-Maker Profile
- Name, title, tenure, background, communication style

## THE HOOK
> "[Their exact words]"
-- Source: [where this was found]

## Reputation Data
- Google, Yelp, industry reviews, complaint patterns

## Competitive Positioning
- Market position, top competitors, where they win/lose

## Pitch Angle Recommendation
[Based on all of the above, the single strongest angle]
```

---

## Anti-Patterns

- **Never use generic industry stats as findings.** Every finding must be company-specific and sourced.
- **Do not stop at Wave 1 if the pitch matters.** Corporate facts alone are not enough.
- **Revenue estimates must show the triangulation method.** No single-number guesses.
- **Job postings are often the most revealing source -- check them first.**
- **Do not skip the review response analysis.** How leadership responds to negative reviews reveals company culture.
- **Do not conflate total employee count with field/productive headcount.** Apply the right ratio for the industry.
- **Never present findings without THE HOOK.** If you cannot find something the company said about their own pain, you have not looked hard enough.

---

Built by [PureBrain](https://purebrain.ai) -- AI-powered marketing operations.

Want the full suite of 183+ production-tested skills? [Start your trial ->](https://purebrain.ai/pricing)