Builder’s Guide to GTM with Apollo
Three paths for the technical GTM builder — API for full control, CLI for speed, MCP for your whole team. Built around a real campaign with 1,297 contacts, open-source scoring, and a cloneable repo.
Everything from the webinar
Every resource from the session, linked from one page.
the waterfall demo as a step-by-step SOP: the commands, the gates, the output
API, CLI, and MCP paths into the same Apollo data
the full kit: chapters, starters, skills
intent-gated lookalike expansion. the list grows itself
Builder's Guide to GTM with Apollo. Aug 13, 2026
every repo on the site, one directory
the data layer everything here runs on
the managed version: campaigns run for you
The campaign behind this
This repo was built around a real outreach campaign. 1,297 GTM engineers sourced from Apollo, segmented into three personas: SaaS operators (850), agency builders (289), and AEO marketers (158). Launched via Smartlead across 44 sending boxes.
Verified emails only, no catch-all domains. That standard, plus a self-hosted LinkedIn layer that puts a personal note on every touch, turned into 272 accepted connections with the exact buyers the campaign targets.
The expansion play takes those 272 companies where we have a relationship and finds the full buying committee at each. That is what the demo shows.
Three paths, same data
Developers building custom pipelines
Python scripts for batch processing, custom scoring, database integrations. Full control over every parameter.
- expand_buying_committee.py — companies to ranked committees
- enterprise_filter.py — the BlackRock problem
- lib/apollo_client.py — clean wrapper with rate limiting
Technical builders who live in the terminal
Shell scripts using Apollo CLI + jq. Composable, pipe-native, cron-ready.
- expand_committee.sh — expansion play, shell-native
- campaign_rerun.sh — rebuild a campaign in shell
- company_intel.sh — enrichment + news + jobs
Reps and non-technical team members
Connect Apollo MCP to Claude. Anyone prospects via chat — no code, no training.
- setup-guide.md — 5-minute connection guide
- example-prompts.md — ready-to-use prompts
The expansion play
You have one contact at a company. The expansion play finds the full buying committee — scored by title relevance and reachability.
# Single company
python3 api/expand_buying_committee.py --domains mixpanel.com
# Batch with Google Sheet output
python3 api/expand_buying_committee.py \
--file data/sample_domains.txt \
--sheet "My Expansion"
# CLI equivalent
./cli/expand_committee.sh --domain mixpanel.com --top 5Scoring system
Each contact gets a composite score: title relevance (additive keywords) multiplied by a reachability tier. A VP of Revenue Operations with verified email and phone scores (100 + 90 + 25) × 1.0 = 215.
| RevOps / Revenue Operations | 100 |
| Revenue | 90 |
| Growth | 80 |
| GTM / Go-To-Market | 75 |
| Sales | 60 |
| Marketing | 55 |
| Product | 50 |
| C-suite | +30 |
| VP | +25 |
| Head of | +20 |
| Director | +15 |
The list grows itself
The waterfall release (v0.8.0) turns a seed list into a full market map with one command. Gates drain from deepest intent to firmographic: Reddit buyer signals, hiring for the pain, fresh funding, tech-stack twins, then lookalikes. Every company lands tagged with the gate it came through, so the sheet explains itself and outreach routes by intent tier.
Full breakdown in the v0.8.0 release notes.
What costs credits (and what does not)
The expansion play scores contacts using free availability flags. It never reveals actual emails or phone numbers. You build the full prioritized list without spending credits, then only reveal the winners.
| Endpoint | Cost | What you get |
|---|---|---|
| mixed_people/api_search | Free | Names, titles, email_status flags, has_phone flags |
| organizations/enrich | Free | Company info: employees, funding, tech stack |
| Email reveal | 1 credit | Actual email address |
| Mobile reveal | Mobile credit | Direct-dial phone number |
Build a ranked list of 100+ contacts using free searches. Pick the top 20 by composite score. Only reveal those 20. That is 20 credits instead of 100+.
Protect your API key
If you are building a repo for your team or sharing code publicly, your API key should never appear in git history.
.env # your real key (gitignored, never committed)
.env.example # template with placeholders (committed)
.gitignore # blocks .env from being trackedThe API client reads from the environment. The test: run git log --all -p | grep APOLLO on your repo. If it returns anything, your key leaked.
The next level is a local secrets vault: one database outside every repo, keys checked out on demand, rotated once, re-pulled everywhere. The pattern is written up in chapter 04 of the gtm-coding-agent.
The repo has everything — API scripts, CLI tools, MCP setup guide, sample data, and a scoring system you can customize for your ICP. Set your Apollo API key and run the expansion play in under 5 minutes.
this whole pipeline is the manual version of what Clearbox runs for you. every gate, every score, every list, managed.
See your market. Move first. →