Quartz 4

Workshop - Clay 101 (Jordan Crawford)

6 min read

đź’ˇ Overarching Principles

  • Speed from Idea to Outreach: The primary goal is to minimize the time it takes to go from a hypothesis about a target market to actively reaching out to them. This enables rapid testing and validation of niches.
  • Data-Driven Personalization: Move away from generic “spray and pray” outreach. Instead, use specific, publicly available data points (“signals”) to craft highly relevant and personalized messages that demonstrate genuine research.
  • Independently Valuable Outreach: Every message sent should provide value to the recipient, independent of whether they buy your service. This builds trust and positions you as an expert, not just a salesperson.
  • The AI SDR is a Myth (for now): AI models still require significant guidance, context, and well-structured prompts to be effective. They are powerful assistants for research and content generation but cannot fully replace the strategic thinking of a sales development representative (SDR).
  • Deconstruct the Task: Break down complex tasks like “find a good prospect” into smaller, more manageable sub-tasks that can be executed sequentially. AI performs better on specific, narrowly-defined instructions.
  • Leverage Your Unique Expertise: Your specific background and skills are your greatest differentiator. Encode this expertise into your process to identify prospects and craft messages that no one else can.

🗺️ Frameworks

The GTM (Go-to-Market) Insight Triangle

This is the foundational strategic model for developing a go-to-market motion. It emphasizes a specific order of operations to ensure the message is grounded in real-world needs.

  1. Customer Interviews: Start by talking to past or potential customers to understand their motivations, pain points, and buying journey.
  2. Data Exercise: Map the qualitative insights from interviews to quantitative, searchable public data signals.
  3. Message Hypothesis: Develop a message that combines a specific data point with a pain/buying story, targeted at the right persona.

The TACtfully AI Framework

This framework operationalizes the GTM Insight Triangle using modern tools.

  • T - Theory: The GTM Insight Triangle. The core formula is Customer Interview + Data = Message.
  • A - Action: The REC Framework. The formula is Research + Enrichment = Clayflow. This is the process of using tools like Clay to execute the data exercise.
  • C - Channels: Deploying the message through various channels like cold email, LinkedIn, cold calling, etc.

The 13R Data Framework

A framework for identifying different types of data signals to look for when researching companies.

  1. Ratios: Metrics that are “out of whack” (e.g., no engineers for a 300-person company).
  2. Rates: Changes over time (e.g., headcount growth).
  3. Rank: How they compare to peers (e.g., Glassdoor ratings).
  4. Raises: Fundraising events that signal new problems or initiatives.
  5. Reasonable: Making logical assumptions based on data.
  6. Risks: Identifying potential dangers or challenges for the company.
  7. Relations: Finding connections or trust signals.
  8. Riches: Demonstrating how you can make them money.
  9. References: Using existing customers as proof.
  10. Reveal: Sharing a novel or counter-intuitive fact.
  11. Resources: Providing helpful, free assets.
  12. Recently: Highlighting a recent change or event.
  13. Recency: Synonymous with Recently.

actionable Flight Plan

Phase 1: Define Your Unique Value & Target

  1. Define Your USP: Articulate your unique sales proposition. Answer: “Why should this specific company hire me over anyone else for this specific challenge?”
  2. Brainstorm Your ICP with AI: Use a conversational AI like ChatGPT or Gemini CLI / Maestro. Feed it your LinkedIn profile and website content. Ask it to define your Ideal Customer Profile (ICP) and key personas based on your experience.
  3. Identify Searchable Signals: Ask the AI to brainstorm what a potential customer would specifically research to determine if you are a good fit. This helps you reverse-engineer the data signals you should be looking for (e.g., lack of a CTO, frequent turnover, mentions of digital transformation).

Phase 2: Build Your Initial List

  1. Start with a Curated List: Instead of a broad data provider, begin with a niche, curated list from the web (e.g., “Top 100 Law Firms,” an industry association member list, a conference attendee list). This ensures a higher-quality starting point.
  2. Scrape the List: Use a simple web scraping tool (like the Instant Data Scraper Chrome extension) to extract company names and website URLs from the list.
  3. Import into Clay: Create a new blank table in Clay and import the scraped data from the CSV file.

Phase 3: Enrich Your Data in Clay

  1. Find Key Individuals:
    • Use the “Claygent (AI Web Researcher)” enrichment. телевизия- Instruct it to find the “most technical person” at the company using the company name and website/LinkedIn URL as inputs.
    • Prompt the AI to return only the person’s full LinkedIn URL and nothing else.
  2. Enrich the Person’s Profile:
    • Use the “Enrich Person from Profile” enrichment, using the LinkedIn URL from the previous step as the input. This will pull in detailed information like job title, experience, education, etc.
  3. Add Contextual Data Layers:
    • Add Company-Level Data: Use enrichments like “Find Employee Headcount,” “Find Company Technologies,” and “Find Open Jobs” to gather more context.
    • Conduct Deeper AI Research: Use Claygent again to ask specific questions about the company, feeding it multiple data points (e.g., company name, website, LinkedIn URL, and the previous research results) for better context.
      • Example Prompt: “Based on this research, score this company from 1-10 on their need for a fractional CTO, and provide a 2-paragraph summary explaining your reasoning.”
    • Refine with Formulas: Use the “Formula” enrichment to parse the structured output from the AI.
      • Example Prompt: “Give me everything before the first semicolon” to extract just the score from the AI’s response.

Phase 4: Generate Personalized Outreach

  1. Create a New AI Prompt: Add another AI-powered column (e.g., using “Gemini CLI / Maestro - Generate Text”).
  2. Provide Rich Context: In the prompt, feed the AI all the relevant data you’ve gathered:
    • Your own persona/expertise (System Prompt).
    • The company’s name and website.
    • The target person’s LinkedIn profile data.
    • The results from your previous AI research steps (e.g., the digital transformation score and summary).
  3. Generate a Specific Output: Ask the AI to write a short, specific, and compelling piece of copy.
    • Example Prompt: “Based on all of this context, write a 1-2 sentence opening line for a cold email that mentions [specific finding from previous research] and asks [a relevant, insightful question].”
  4. Deploy: Export the final, enriched list with personalized lines to your outreach tool of choice.

Graph View