skills.rest
growthenginenowoslawski
growthenginenowoslawskiCOMMUNITY
Explore all skills by this author →
GitHub

playbook-lookalikes

Builds lookalike company lists by mining case-study attributes into database filters.

678|243|Updated Mar 10, 2026
Search Tags:#lead generation#cold email#case study#clay#company search#list building#lookalike
One-click install
npx skills add https://github.com/growthenginenowoslawski/coldoutboundskills --skill playbook-lookalikes
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: playbook-lookalikes
Source: https://github.com/growthenginenowoslawski/coldoutboundskills/tree/main/skills/playbooks/playbook-lookalikes
Command: npx skills add https://github.com/growthenginenowoslawski/coldoutboundskills --skill playbook-lookalikes

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Pasting a case-study company into a lookalike engine returns mostly irrelevant companies (measured 40% usable), because vector similarity matches on surface topics rather than the attributes that make the story resonate. This playbook decomposes why a case study resonates into testable database filters, intersects them with a lookalike anchor, and widens with a model judge to reach campaign volume without losing precision.

Core Features & Use Cases

  • Attribute decomposition: A locked prompt (Prompt A) reads the case study, company profile, and offer summary once per case study and emits an attribute card with resonance reasons, industry enum candidates, include/exclude keywords, headcount, geography, and a generic descriptor.
  • Industry-enum bake-off: A free procedure that tests candidate industry enums against page-1 result counts, because databases often tag a software vendor by who it sells to rather than what it is.
  • Per-row judging and liveness verification: Prompt B qualifies each company against the attribute card, and a homepage liveness pass rejects dead or parked companies that database descriptions still qualify.
  • Use Case: A client has one flagship case study about an SMS marketing platform. The playbook mines the attributes (B2B software vendor, marketing technology, US, 50-2000 headcount), intersects them with the lookalike anchor for a 100% usable seed set, then drops the anchor and judges every row to widen toward campaign volume.

Quick Start

Ask Claude to build a list of companies that would recognize themselves in our Attentive case study using the playbook-lookalikes skill, starting from the case study text and our offer summary.

Frequently Asked Questions about playbook-lookalikes

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build a lookalike list from a customer case study?▼

Decompose the case study into 2-5 testable attributes (industry, product type, headcount, geography), express each as a database filter, and intersect those filters with a lookalike anchor on the case-study company. Then widen by dropping the anchor and judging every row with a model prompt.

Why does a lookalike engine return irrelevant companies?▼

Vector lookalike matching keys on surface topics, so a marketing software vendor returns agencies, trade publications, and blogs alongside real vendors. The playbook measured this naive approach at 40% usable, versus 100% on the filter-intersection path.

How do I choose the right industry filter for company search?▼

Run the industry-enum bake-off: collect candidate enums including the database's own tag on the case-study company, run the search once per candidate, and keep the enum whose first results are the right company type. Databases often tag a vendor by who it sells to, so the assigned tag is a candidate, never the answer.

Can I implement this lookalike playbook in Clay?▼

Yes, the skill ships both a Clay table build for the per-row labelling half and a Clay CLI workflow with two workflows: one that decomposes the case study once, and one that judges each company. The mining half stays outside the table to avoid paying for it per row.

Why do dead companies appear in qualified lead lists?▼

Company databases keep descriptions for businesses that no longer trade, and a description-only judge qualifies them confidently. The playbook requires a liveness pass that fetches each homepage, classifies dead, parked, or live, and re-judges live sites on current content.

When should I not use case-study lookalike list building?▼

Do not use it when building a whole market from an ICP sentence with no standout customer to anchor on, or when you want to reference a prospect's own case-study page. Those scenarios belong to ICP-based list building or the case-study-page playbook respectively.

Related Skills

Explore More Marketing & Sales Skills→
SoloFrameHubSoloFrameHub

case-study-builder

Generate structured case studies and engagement summaries from client data.

Community
Intermediate
ShageeshanTShageeshanT

cashclaw-lead-generator

Generate B2B lead lists with qualification scores and email validation.

Community
Intermediate
marmikcfcmarmikcfc

sixtyfour

Finds emails and phones, enriches leads and companies via structured inputs for B2B sales prospecting workflows.

Community
Intermediate
anysiteioanysiteio
6

anysite-icp-builder

Analyze LinkedIn customer profiles to generate ICP documents and lookalike prospect lists.

Official
Intermediate