GTM Strategy Generator

Drop your URL. Get 9 custom outbound plays — each with full execution playbook, AI Ark + Apollo filters, and 2 email variants — as JSON. Free CLI from BuzzLead.

The same engine that powers the free web tool — open-sourced for operators who want to run it on their own keys.

npx @buzzlead/gtm-strategy --url yourcompany.com --out strategy.json

In ~2-3 minutes you get a JSON file with 9 distinct outbound plays for your business, each carrying:

• The angle, type, difficulty, and expected response rate tier
• A full execution playbook (5-7 steps + why it works)
• Ready-to-paste AI Ark filter values (titles, exclusions, employee range, industries, keywords, seniorities, geos, signals)
• Ready-to-paste Apollo filter values (same shape, Apollo conventions)
• Estimated list size
• 2 email variants per play (with 3 subject line options + follow-up + framework explanation)

What it does

Five steps, fully automated:

1. Reads your site — Jina Reader pulls clean markdown, Claude extracts your offer, ICP, differentiators, industry, target titles.
2. Builds market intel — Exa surfaces industry trends, competitor moves, and Reddit complaint threads in your category. Claude synthesizes them.
3. Designs 9 plays — Signal / ICP / Competitor / Trigger / Pain / Timing / Account-Based. Difficulty + response-rate tiers balanced.
4. Expands each play into a playbook — full execution steps, ready-to-paste AI Ark + Apollo filter fields, list size estimate.
5. Writes 2 cold email variants per play — picked from the right framework for each play type (4T, Pain Amplification, Authority Opener, Trojan Horse, AIDA).

Install

One-shot via npx (no install)

npx @buzzlead/gtm-strategy --url yourcompany.com

Or install globally

npm install -g @buzzlead/gtm-strategy
gtm-strategy --url yourcompany.com

Or use as a library

import { runStrategy } from "@buzzlead/gtm-strategy";

const result = await runStrategy({
  url: "yourcompany.com",
  onProgress: (step) => console.log(step.kind, step.status),
});

console.log(result.plays); // 9 plays, each with playbook + copy

API keys

Two keys required.

| Key | Why | Where | |---|---|---| | ANTHROPIC_API_KEY | Site analysis, research synthesis, play generation, playbook + copy | https://console.anthropic.com | | EXA_API_KEY | Industry trends + competitor moves + Reddit complaints | https://exa.ai |

Optional:

• JINA_API_KEY — faster site scrape (anonymous works)

Copy .env.example to .env and fill them in.

Cost per run

Approximate, full 9-play expansion:

| Provider | Calls | ~Cost | |---|---|---| | Anthropic (sonnet-4-6) | 1 + 1 + 1 + 9 + 9 = 21 calls | $0.50–1.20 | | Exa | 3 searches | ~$0.02 | | Jina | 1 scrape | ~$0.001 | | Total | | ~$0.50–1.20 |

For a faster, cheaper preview, use --plays-only (drops to ~$0.05 and ~30 seconds — gets the 9 angles without playbook or copy).

CLI usage

Usage: gtm-strategy [options]

Required:
  -u, --url <url>            Your company URL

Optional:
  -n, --name <name>          Company name hint (overrides URL-derived default)
      --plays-only           Skip playbook + copy expansion (fast preview)
      --expand <n>           Only expand the first N plays (saves cost)
  -o, --out <path>           Output JSON path. Default: ./gtm-strategy.json
      --stdout               Write JSON to stdout instead of file
      --silent               Suppress progress output
  -h, --help                 Display help
  -V, --version              Display version

Examples

Full 9-play expansion:

gtm-strategy --url stripe.com --out stripe-gtm.json

Fast preview (just the 9 angles, no playbooks/copy):

gtm-strategy --url stripe.com --plays-only --out stripe-preview.json

Expand only the top 3 plays:

gtm-strategy --url stripe.com --expand 3 --out stripe-top3.json

Pipe into another tool:

gtm-strategy --url stripe.com --plays-only --stdout | jq '.plays[].name'

Output schema

The JSON file has this top-level shape:

{
  companyIntel: {
    companyName, websiteUrl, offer, icp,
    differentiators[], industryVertical, targetTitles[]
  },
  researchIntel: {
    industryTrends[], competitorMoves[],
    redditComplaints[], macroContext
  },
  plays: [
    {
      id, name, type, angle, difficulty, responseRateTier,
      playbook: {
        fullDescription, messagingAngle, steps[],
        aiArkFields: { jobTitles[], jobTitlesExclude[], companySizeMin/Max,
                       industries[], keywords[], seniorityLevels[],
                       geographies[], signals[] },
        apolloFields: { personTitles[], excludedTitles[], employeeRanges[],
                        industries[], keywords[], seniorityLevels[], locations[] },
        estimatedListSize, whyItWorks
      },
      copy: {
        variant1: { framework, subjectLines[], emailBody, followUp, whyItWorks },
        variant2: { framework, subjectLines[], emailBody, followUp, whyItWorks }
      }
    },
    // ... 8 more
  ]
}

How it differs from the web tool

| | Web tool | CLI | |---|---|---| | Plays per run | 9 | 9 | | Lead gate | Required | None | | Auth | Shared keys | Your keys | | Cost | Free | Your API spend (~$1) | | Output | Web UI + email copy buttons | JSON file (or stdout) | | Plays-only mode | No | Yes (--plays-only) | | Library API | No | Yes |

Library API

import {
  runStrategy,        // end-to-end pipeline
  analyzeSite,        // step 1: extract CompanyIntel from a URL
  buildResearch,      // step 2: market research via Exa + Claude
  generatePlays,      // step 3: generate 9 plays
  generatePlaybook,   // step 4: per-play playbook expansion
  generateCopy,       // step 5: per-play 2-variant copy
} from "@buzzlead/gtm-strategy";

Each step is a pure async function — swap, compose, or run individually.

Want this run for an entire pipeline?

GTM Strategy Generator is a single-shot tool. It gives you the strategy. You take it from there.

BuzzLead the agency runs this kind of GTM motion continuously — strategy, list building, infrastructure, copy, sending, reply handling. 10M+ cold emails sent. $8M+ in client revenue.

If you want the version with humans behind it, talk to us at buzzlead.io.

License

MIT. Use it however you want.