Skip to content
Back to skills

Prospect Audit

ASecurity

Generate full prospect audit as a sales tool before demos. Combines company research, CRO audit, and person intel into a presentable document with talking points and opportunity sizing.

  • 2 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added September 10, 2026
content-marketingrustgo

Works with

  • cli

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned September 10, 2026

npx -y skills add ekatasingh1107/b2b-gtm-skills --skill prospect-audit --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Prospect Audit?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Prospect Audit
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/ekatasingh1107-prospect-audit/badge)](https://www.skillsdirectory.com/skills/ekatasingh1107-prospect-audit)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: prospect-audit
description: >
  Generate full prospect audit as a sales tool before demos. Combines
  company research, CRO audit, and person intel into a presentable
  document with talking points and opportunity sizing.
tags: [demo-prep, audit, sales-tool, playbook]
---

# Prospect Audit

Pre-demo sales preparation tool. Generates a comprehensive audit document for a specific prospect, combining company intelligence, CRO findings, and decision-maker research into a conversation-ready deliverable. Use this before every demo to walk in with authority.

## Prerequisites

- `agency.config.json` fully populated with case studies and service offerings
- Tool stack configured (Apollo for person research, web access for CRO audit)
- Company website must be live and accessible

## Composites & Capabilities Used

1. `company-researcher` -- deep company intelligence gathering
2. `cro-auditor` -- full store audit across all page types
3. `person-researcher` -- decision-maker background and activity
4. `decision-maker-finder` -- identify additional stakeholders

## Phase 0: Intake

Accept the following inputs:
- **Company name**: the prospect's business name
- **Website URL**: their primary website (Shopify store preferred)
- **Contact name**: the person the demo is scheduled with
- **Demo date**: when the demo is happening

Validate that the website is accessible before proceeding. If the site is down or behind a password wall, flag immediately and proceed with company research only.

## Phase 1: Company Deep Dive

Execute `company-researcher` with depth: "deep"

Gather:
- Business overview: what they sell, their positioning, target market
- Tech stack: Shopify plan, theme, apps installed, payment providers, analytics tools
- Social presence: Instagram followers, engagement rate, LinkedIn company page, Twitter/X
- Recent news: funding rounds, product launches, leadership changes, press mentions
- Growth signals: hiring activity, new market expansion, ad spend changes
- Competitive landscape: who are their direct competitors, how do they compare
- Reviews and reputation: Trustpilot, Google reviews, app store reviews if applicable

## Phase 2: CRO Audit

Execute `cro-auditor` with depth: "deep"

Audit the following areas:
- **Homepage**: hero, value proposition, trust signals, navigation, above-the-fold CTA
- **Product pages**: imagery, descriptions, variant selection, add-to-cart UX, social proof
- **Collection pages**: filtering, sorting, product card design, pagination
- **Mobile experience**: responsive behavior, thumb-zone CTA placement, load time
- **Cart and checkout**: cart drawer vs page, upsells, trust badges, shipping clarity
- **Site speed**: Core Web Vitals, image optimization, script bloat
- **SEO basics**: meta titles, descriptions, structured data, internal linking

Score each area out of 10. Calculate an overall score. Rank findings by impact (revenue at stake) and effort (implementation complexity).

## Phase 3: Person Research

Execute `person-researcher` on the demo contact:
- LinkedIn profile summary and headline
- Recent LinkedIn posts and articles (last 3 months)
- Conference talks, podcast appearances, media quotes
- Career trajectory: previous roles, time at current company, promotions
- Topics they engage with: what they like, comment on, share
- Mutual connections or shared interests
- Communication style indicators (formal vs casual, data-driven vs narrative)

## Phase 4: Compile Audit Document

Structure the output as a presentable audit:

```
PROSPECT AUDIT: [Company Name]
Prepared for: [Contact Name], [Title]
Date: [Demo Date]
Prepared by: Plasho

=== EXECUTIVE SUMMARY ===
[2-3 sentence overview: what the company does, what we found, what's at stake]

=== COMPANY OVERVIEW ===
- Business: [summary]
- Platform: [tech stack]
- Team: [size if discoverable]
- Recent news: [funding, launches, hires]
- Social footprint: [Instagram Nk, LinkedIn Nk]

=== STORE AUDIT FINDINGS ===
Overall Score: X/10

Finding 1: [Issue Title]
- What: [Description of the issue]
- Impact: [What it's costing them -- quantified where possible]
- Fix: [Specific recommendation]
- Effort: [Low / Medium / High]
- Priority: [P1 / P2 / P3]

Finding 2: [Issue Title]
- What: [Description]
- Impact: [Cost]
- Fix: [Recommendation]
- Effort: [Level]
- Priority: [Level]

[Continue for top 5-7 findings, ordered by impact]

=== STRENGTHS ===
- [What they're doing well -- always start the conversation positive]
- [Specific things to compliment]
- [Areas where they're ahead of competitors]

=== OPPORTUNITY SIZING ===
Based on industry benchmarks:
- Current estimated conversion rate: X%
- Potential with fixes: Y%
- Revenue impact: Z% increase in online revenue
- If monthly revenue is ~$A, the gap is ~$B/month

=== RELEVANT CASE STUDY ===
[Most relevant case study from agency.config.json case_studies]
- Client: [name]
- Challenge: [what they faced]
- Solution: [what was done]
- Result: [metrics and outcomes]
- Relevance: [why this matters to the prospect]

=== TALKING POINTS FOR DEMO ===
1. [Personalized opening based on person-researcher -- reference their recent post or talk]
2. [Key finding to lead with -- the highest-impact issue from the audit]
3. [Discovery question to ask them -- something that shows you've done homework]
4. [Case study to reference -- tie it to their specific situation]
5. [CTA / next step to propose -- clear, specific, time-bound]
```

## Phase 5: Output

Return the audit in two formats:
1. **Structured text**: the full document as formatted above, ready to read or paste
2. **JSON**: machine-readable version with all fields for programmatic use or CRM storage

```json
{
  "company": {
    "name": "",
    "website": "",
    "industry": "",
    "tech_stack": [],
    "social": {},
    "news": []
  },
  "contact": {
    "name": "",
    "title": "",
    "linkedin": "",
    "talking_points": []
  },
  "audit": {
    "overall_score": 0,
    "findings": [],
    "strengths": [],
    "opportunity": {}
  },
  "demo_prep": {
    "opening": "",
    "key_finding": "",
    "discovery_question": "",
    "case_study": "",
    "cta": ""
  }
}
```

Optionally: save the audit to CRM notes for the lead, and export as HTML for sharing with the team.

## Example Usage

**Trigger phrases:**
- "Audit [company] before my demo"
- "Prep for demo with [contact] at [company]"
- "Run prospect audit for [website]"
- "Generate sales prep for tomorrow's demo"

Files in this skill

  • SKILL.md6.3 KB
  • skill.meta.json515 B

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.