Lead Quality Diagnosis

---
name: lead-quality-diagnosis
description: Find why a campaign is producing leads that do not turn into sales conversations, pipeline, or revenue.
---

# Lead Quality Diagnosis

## What This Skill Does
Find why a campaign is producing leads that do not turn into sales conversations, pipeline, or revenue.

Expected outcome: A lead-quality diagnosis across targeting, offer, form design, qualification, and sales handoff.

## When To Trigger
Use this skill when the user needs a practical marketing diagnosis, plan, audit, or reusable workflow related to:
- B2B services
- SaaS
- high-ticket lead generation
- sales-led teams

Do not use this skill for generic brainstorming. Use it when there is a real marketing problem, dataset, page, campaign, channel, brief, or decision to inspect.

## Compatible Agents
- Claude Code
- Codex
- Agent environments that can read Markdown skill instructions
- Scripted workflows that pass inputs as files, URLs, JSON, CSV, screenshots, or plain-text notes

## What The Agent Should Ask For
Ask for missing inputs before doing the final analysis. If the user wants a partial run, continue only after marking every assumption clearly.

Required or strongly recommended inputs:
- Lead export: provide a file path, URL, export, screenshot, or short notes.
- campaign source: provide a file path, URL, export, screenshot, or short notes.
- form fields: provide a file path, URL, export, screenshot, or short notes.
- CRM stages: provide a file path, URL, export, screenshot, or short notes.
- sales feedback: provide a file path, URL, export, screenshot, or short notes.

Useful context:
- Business goal and target customer
- Date range or campaign period
- Channel, platform, market, and budget context
- Current hypothesis or decision the user needs to make
- Any constraints, such as team capacity, tracking limits, brand rules, or compliance concerns

Ask for missing inputs with short, specific questions. Do not invent missing data, conversion rates, spend, audience settings, CRM outcomes, or performance history.

## How To Run
1. Restate the goal in one sentence and list the inputs received.
2. Identify missing information and decide whether to ask for it or proceed with labeled assumptions.
3. Normalize the evidence into comparable units, such as date range, channel, campaign, page, audience, offer, metric, or lifecycle stage.
4. Run the skill-specific workflow below.
5. Separate facts from assumptions and confidence levels.
6. Rank every recommendation by impact, confidence, and effort.
7. Produce an action plan that a marketer, founder, or agent can execute without needing to reverse-engineer your thinking.

## Skill-Specific Workflow
1. Trace leads from click to outcome
2. Identify fit and intent problems
3. Review offer and form design
4. Recommend qualification changes

## Outputs
- Bad-fit lead patterns: include evidence, recommendation, owner, and next action.
- qualification fixes: include evidence, recommendation, owner, and next action.
- offer recommendations: include evidence, recommendation, owner, and next action.
- handoff checklist: include evidence, recommendation, owner, and next action.

## Output Format
Return Markdown in this structure:

# Lead Quality Diagnosis Report

## Executive Summary
- 3 to 5 bullets on what matters most.

## Evidence Reviewed
- List the files, URLs, exports, screenshots, notes, and assumptions used.

## Findings
Use a table with these columns:
| Finding | Evidence | Impact | Confidence | Effort | Owner |

## Recommendations
Rank recommendations from highest to lowest priority. For each recommendation include:
- What to change
- Why it matters
- How to implement it
- How to measure success

## 30-Day Action Plan
- Week 1
- Week 2
- Week 3
- Week 4

## Missing Data
- List anything that would change the conclusion if supplied.

## Script/Automation Notes
- Include file paths, generated artifacts, machine-readable summaries, or follow-up commands when relevant.

## Quality Bar
- Every finding must cite evidence from the supplied inputs or be labeled as an assumption.
- Do not invent missing data.
- Recommendations must be ranked by impact, confidence, and effort.
- Separate immediate fixes from experiments.
- Keep the answer specific enough that another agent can turn it into tasks.
- Prefer a smaller number of strong, evidence-backed recommendations over a long generic checklist.
- If the evidence is weak, say so plainly and explain what would make the analysis stronger.

## Guardrails
- Do not log into ad platforms, analytics tools, CRMs, CMSs, or customer systems unless the user explicitly asks.
- Do not change campaigns, budgets, tags, tracking, website code, CRM fields, email flows, or live content unless the user explicitly asks.
- Do not expose private customer data in the final output.
- Do not present guesses as facts.
- If the task touches regulated, sensitive, or legally risky claims, flag the risk and keep recommendations conservative.

## Script/Automation Use
This skill can be called from a script or agent runner by passing:
- skill: lead-quality-diagnosis
- goal: the business question to answer
- inputs: file paths, URLs, exports, screenshots, or notes
- dateRange: the period covered by the evidence
- outputPath: where the Markdown report or structured artifacts should be written

Return Markdown for humans and, when useful, a JSON-ready summary with:
- findings
- recommendations
- missingData
- actionItems
- confidence

Expected artifacts:
- Bad-fit lead patterns
- qualification fixes
- offer recommendations
- handoff checklist

## Handoff Prompt
Use this prompt when triggering the skill in Claude Code, Codex, or another agent:

You are running the Wieldr Lead Quality Diagnosis skill.

Goal: [state the business question]
Inputs: [paste notes or list file paths/URLs]
Date range: [state period]
Output destination: [state where the report should go]

Follow the SKILL.md exactly. Ask for missing inputs if they are required. If asked to proceed with incomplete data, label assumptions clearly. Return Markdown using the required output format. Rank every recommendation by impact, confidence, and effort.

## Source
Canonical page: https://wieldr.com/skills/lead-quality-diagnosis
Download URL: https://wieldr.com/skills/lead-quality-diagnosis/SKILL.md