SourcingICP · PersonaQualification · Scoring

Niche Signal Discovery

Discovers differential signals between Closed Won and Closed Lost accounts by scraping multi-page website content and job listings, then computing Laplace-sm...

by Aero·MIT·18 on GitHub·updated 2026-05-15
npx skills add getaero-io/gtm-eng-skills

Description

Discovers differential signals between Closed Won and Closed Lost accounts by scraping multi-page website content and job listings, then computing Laplace-smoothed lift scores. Identifies what observable pre-sales signals — job listings, website language, tech stack, compliance infrastructure — distinguish buyers from non-buyers, and outputs 10 net-new prospects matching the winning signal profile.

What this skill does

  1. 1Discovers the target company's product, buyer persona, and competitive ecosystem before generating any signals.
  2. 2Scrapes multi-page website content and job listings for both Closed Won and Closed Lost account lists using Serper and Firecrawl.
  3. 3Runs Laplace-smoothed differential analysis to compute which signals appear significantly more often in Won accounts than Lost ones.
  4. 4Generates a structured report with evidence-backed signals, lift scores, and cited quotes — every top signal must have three or more source examples.
  5. 5Outputs 10 net-new prospects that match the winning signal profile, with optional contact and email discovery.

When to use

  • You have a list of closed won and closed lost accounts and want to know what actually differentiates buyers from non-buyers.
  • You are building a lead scoring model and need first-party signals — job listings, website language, tech stack — not CRM-derived fields.
  • Your ICP definition is vague and you want to ground it in observable pre-sales signals with statistical backing.
  • You want to deliver a signal-based prospect list alongside an ICP analysis report.

Tools used

Apollo· EnrichmentSlack· Productivity

Best for

Revenue OperationsGTM EngineerSales Development RepFounder / CEO

Format

FrameworkAnalyzerPlaybook

Frequently asked questions

What is Niche Signal Discovery?

Niche Signal Discovery is a Claude Code skill for sourcing by Aero. Discovers differential signals between Closed Won and Closed Lost accounts by scraping multi-page website content and job listings, then computing Laplace-sm...

When should I use Niche Signal Discovery?

You have a list of closed won and closed lost accounts and want to know what actually differentiates buyers from non-buyers. You are building a lead scoring model and need first-party signals — job listings, website language, tech stack — not CRM-derived fields.

How do I install Niche Signal Discovery?

Run the install command (npx skills add getaero-io/gtm-eng-skills). Once the skill is in place, Claude Code loads it automatically whenever a task matches what it does.

Is Niche Signal Discovery free to use?

Yes. Niche Signal Discovery is open source under the MIT license, published in the getaero-io/gtm-eng-skills repository on GitHub.

Source repository

More from getaero-io/gtm-eng-skills

Other skills that might interest you