muzzdang

Second Brain notes from YouTube & builds

Second Brain notes

Jev: I Used Jev To Build a List of 10,000 Ready-to-Buy Leads From Scratch in 12 Minutes

Watch on YouTube: https://www.youtube.com/watch?v=H8PKEb7dYTs

Channel: Aaron Shepherd | GrowthFlare

Building a 10,000-Lead Cold Email List in Under 30 Minutes: A Practical Breakdown

Most cold outreach workflows stall at the data layer — either because per-credit pricing makes broad pulls expensive, or because manual qualification eats hours. In a recent walkthrough, Aaron Shepherd of GrowthFlare demonstrates an end-to-end pipeline that moves from raw company data to 16,902 persona-segmented, ICP-qualified contacts in roughly 30 minutes. The stack combines flat-fee data providers, waterfall email enrichment, and a decision-only AI model called Jev to filter at both company and prospect levels. Total variable AI cost: $0.94.

What the Video Covers

The demonstration walks through three distinct stages: (1) pulling a broad initial list of SaaS companies and contacts using GetLead, (2) running a second enrichment pass with Quick Enrich to recover missing emails, and (3) applying Jev — a lightweight classification model — to disqualify non-ICP companies and prospects before loading the remainder into persona-specific campaigns. The emphasis is on replacing per-lead credit systems with unlimited flat-fee subscriptions, then using cheap, high-throughput AI decisions to narrow the funnel.

Step-by-Step Walkthrough

  1. Broad company pull. Using GetLead (unlimited data, flat monthly fee, 82 fields per lead), Shepherd queries for SaaS companies with 20+ headcount. The initial pull returns 10,382 companies and 21,598 contacts across 7,862 companies.
  2. Waterfall email enrichment. Quick Enrich ($99/month for unlimited web app lookups; higher for API) runs against the 2,520 companies that lacked emails after the first pass. It adds 1,344 contacts at 780 companies, lifting overall email coverage from 78% to 83%.
  3. Company-level ICP qualification with Jev. Jev evaluates each company against a SaaS-only criterion. Of ~10,300 companies, 7,700 are classified as likely SaaS, 1,500 as not SaaS, and ~1,000 flagged for review. Cost: $0.27 (4¢ per million tokens).
  4. Prospect-level verification. Jev then checks each contact at qualified companies for three signals: still at the company, holds a buying-role title, and has sufficient scope (not a narrow-scope leader). Of 18,565 prospects, 16,903 pass across 6,442 companies. Cost: $0.67.
  5. Persona segmentation and offer mapping. Qualified leads are grouped into four personas — Founder/CEO, Sales, Marketing, COO — with one tailored offer variation written per segment. The final list of 16,902 leads loads into persona-specific campaigns.

Tools, Numbers, and Unit Economics

  • GetLead: Flat monthly fee, unlimited pulls, 82 fields per lead. Used for the initial broad query.
  • Quick Enrich: $99/month unlimited web app lookups; API pricing higher. Serves as the second enrichment layer.
  • Jev: Decision-only model (yes/no/score, no text generation). ~4¢ per million tokens, sub-0.5-second response. Company check: $0.27. Prospect check: $0.67. Total AI spend: $0.94.
  • Claude Code: Used to script API calls across GetLead and Quick Enrich, automating the pull-enrich-qualify chain.
  • Raw-to-qualified funnel: 34,000 raw pulls → 21,598 initial contacts → 16,902 qualified leads across 6,442 companies.
  • Disqualification breakdown: 60 contacts left the company, 421 held non-buying roles, 514 were narrow-scope leaders, plus non-SaaS companies removed at company level.
  • Fixed costs: “A few hundred dollars” for unlimited GetLead and Quick Enrich subscriptions.

Caveats and Hype Check

Pulling a massive list is easier and cheaper than ever, you have to be really good at narrowing them down to people who would actually buy from you.

Cheap data is not junk data. It’s just that checking every lead against your ICP means hours of someone clicking through websites and LinkedIn profiles.

Several claims warrant scrutiny. The title cites “12 minutes” but the narrated process takes “just over 30 minutes.” Jev is described as a “brand new AI model most people haven’t heard of” — the transcript says “Jeff” while the title says “Jev,” and no public model card or documentation is referenced. The “ready-to-buy” framing conflates ICP fit with purchase intent; qualification confirms role and company type, not buying signals. The video also promotes the creator’s done-for-you agency (“book a call with us using the first link below”) and cites an unnamed finance client who “closed a deal worth more than they’ll ever pay us” — a revenue claim without verifiable details.

Concrete Next Actions

  • Replace per-lead data providers with flat-fee tools (GetLead, Quick Enrich) to pull broad initial lists without credit rationing.
  • Implement waterfall enrichment: run a second data source on contacts missing emails from the first source.
  • Use a decision-only AI model (Jev or equivalent) to qualify companies against ICP criteria before prospect-level checks.
  • Run prospect-level verification (company tenure, buying authority, role scope) with the same low-cost model.
  • Segment qualified leads into 3–5 key personas and write one high-converting offer per persona instead of per-lead customization.
  • Automate the pipeline with Claude Code or similar to chain API calls across data providers and qualification steps.

Personalization doesn’t have to mean a custom first line for every single person. It could also simply mean the right line for each type of lead.

The core takeaway: cheap, unlimited data plus sub-penny AI classification changes the economics of list building. The bottleneck shifts from access to judgment — and that judgment can now be automated at scale.