Niche Signal Discovery
Find first-party website and hiring signals that separate Closed Won from Closed Lost accounts so you can score prospects and tighten ICP criteria.
Overview
niche-signal-discovery is an agent skill most often used in Idea (also Validate scope, Grow lifecycle) that uses Deepline enrichment to discover Laplace-smoothed differential signals between Closed Won and Closed Lost ac
Install
npx skills add https://github.com/code.deepline.com --skill niche-signal-discoveryWhat is this skill?
- Computes Laplace-smoothed lift scores between Won and Lost cohorts
- Enrichment pipeline via `deepline enrich` (serper, firecrawl, crustdata)—no scattered API keys
- Shipped scripts use Python 3 stdlib only—no pip dependencies
- Documented ~0.47 credits per company before optional Step 7 contact discovery
- Triggers on ICP analysis, niche signals, won vs lost, lead scoring, first-party buyer signals
- ~0.47 credits per company for core enrichment (serper 0.02 + firecrawl 0.05 + crustdata 0.40)
- Python 3 stdlib only for shipped scripts—no pip dependencies
Adoption & trust: 4.8k installs on skills.sh; 1/1 security scanners passed (skills.sh audits).
What problem does it solve?
You have won and lost domain lists but no quantitative first-party signals explaining which accounts are worth pursuing.
Who is it for?
Founders doing bottoms-up ICP work with explicit won/lost cohorts who will approve ~0.47 credits/company enrichment via Deepline.
Skip if: Builders without customer outcome labels, those skipping Deepline setup, or anyone expecting zero-cost automated prospecting without user approval on paid steps.
When should I use this skill?
User provides won/lost customer domain lists and wants ICP niche signals, differential analysis, signal discovery reports, or account/lead scoring inputs; read deepline-gtm first.
What do I get? / Deliverables
You get an ICP-oriented signal report with lift-ranked website, hiring, and maturity markers suitable for account scoring and prospecting criteria—after approved Deepline credit spend.
- ICP signal report with lift-ranked differentiators
- Prospecting/scoring criteria derived from won vs lost patterns
- Enriched company context (site content, jobs, stack cues) via Deepline
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
Canonical shelf is Idea because differential buyer signals answer “who actually buys?” before you commit product and GTM bets. Research fits systematic won/lost enrichment and lift scoring rather than building the scoring product itself.
Where it fits
Upload won/lost domains from a spreadsheet export and rank what content and hiring patterns correlate with wins.
Trim a broad TAM hypothesis to segments that show the strongest lift signals before building a scoring MVP.
Refresh prospecting filters quarterly when win/loss mix shifts and job-posting signals change.
How it compares
Use for differential won/lost signal science via Deepline—not a generic web scraper or one-off LinkedIn lookup.
Common Questions / FAQ
Who is niche-signal-discovery for?
Solo builders and lean GTM operators with Closed Won and Closed Lost domain lists who want data-backed ICP and scoring signals through Deepline.
When should I use niche-signal-discovery?
During idea-stage market research when comparing buyers vs non-buyers, validate when narrowing ICP scope from CRM exports, or grow when refreshing prospecting criteria from recent win/loss patterns.
Is niche-signal-discovery safe to install?
It orchestrates paid third-party enrichment through Deepline—always confirm credits and data handling; review the Security Audits panel on this Prism page before running on sensitive account lists.
Workflow Chain
Requires first: deepline gtm
SKILL.md
READMESKILL.md - Niche Signal Discovery
# Niche Signal Discovery Discover differential signals between Closed Won and Closed Lost accounts by extracting multi-page website content and job listings, then computing Laplace-smoothed lift scores to identify what distinguishes buyers from non-buyers. ## Prerequisites - **Deepline CLI** — All enrichment runs through `deepline enrich`. No separate API keys for exa/crustdata/apollo etc. - **Python 3** stdlib only — no pip dependencies for any shipped script. - **Credits** — ~0.47 credits/company (serper 0.02 + firecrawl 0.05 + crustdata 0.40). Step 7 contact discovery is additional. **Always get user approval before paid steps.** ## Deepline-First Principle Use `deepline enrich` for all enrichment, `deepline tools execute` for one-offs, `deepline playground` for inspection. Reruns are idempotent. Refer to `deepline-gtm` for command patterns and provider playbooks. ## Input requirements - Won and lost customer domain lists (≥20 won + ≥10 lost for statistical significance) - **Lookalikes can supplement Won** if Closed Won < 15. Add a Dataset Caveat to the report. - **Target company context** from Step 0 — what they sell, who they sell to, key personas. ## Pipeline ``` 0. Discover target company (what they sell, who they sell to) 0.5. Discover ecosystem (competitors, tech stack, buyer personas) 1. Prepare input CSV (deduplicate within won/lost groups) 1.0.5 Build "do not re-contact" index from user's existing list (scripts/dedupe_utils.py) 1.5. Generate vertical-specific configs (keywords, tools, job roles) 2. Multi-page website + job extraction (deepline enrich) 3. Quality gate — verify file completeness + coverage (>80%) 3.5. Review configs against enriched data 4. Differential analysis (scripts/analyze_signals.py) 5. Generate report — every top signal must include cited evidence 6. Signal interpretation review 7. Top 10 net-new prospects [REQUIRED] + contacts/emails [optional, costs credits] ``` **Step 7 is required.** A signal report without 10 actionable companies forces the reader to do their own prospecting pass — exactly the expensive thing they wanted to skip. Contacts/emails are optional only because they cost extra credits; always offer them. ## Signal reliability hierarchy Highest → lowest confidence: 1. **Job listings** — active budget + acknowledged pain. Highest-intent. 2. **Analyst validation** (Gartner/Forrester) — typically 4-7x lift, rare in lost. 3. **Compliance infrastructure** (SOC2/GDPR/ISO) — procurement maturity. 4. **Buyer pain language** on careers/blog — operational awareness. 5. **Tech stack tools** (niche SaaS) — infrastructure readiness. 6. **Website product/marketing content** — variable; can be buyer OR competitor. **When website signals fail:** For B2B back-office tools (AR, billing, compliance), buyers don't publish their pain on marketing pages. Prioritize jobs + tech stack + firmographics for these verticals. ## What NOT to use for scoring CRM fields populated by AE activity — catalyst note count, OCR-derived counts (`number_of_champions_c`, `number_of_decision_makers_c`), MEDDPICC picklists, any "did the AE do X on this opp" field — correlate with win-rate as **engagement artifacts, not causal signals**. They get fi