19 GTM prompts. Copy and run.
Every prompt I use across my GTM stack — cold email, BANT extraction, account research, reply classification, Clay columns, and n8n automation. Tested in production. Drop them into Claude, GPT-4o, or your Clay AI columns.
Paste in company data and get a structured ICP fit score with reasoning. I use this in Clay to triage lists before they hit any sequence.
Runs as an AI column in Clay. Reads company data from multiple enrichment sources and returns a fit score, tier, and outreach angle per row — so you only write to the top accounts.
Paste all the enrichment signals you have on a prospect and get back a ranked top 5 with outreach hooks. Useful when Clay returns 20+ data points and you need to decide what to lead with.
Paste a job description and extract what's broken in their GTM stack. Job ads are the most honest thing companies publish — they list exactly what's not working.
A company just raised. Here's how I write the outreach — specifically for the 30-day window after a funding announcement when they are actively spending on growth infrastructure.
Generates a personalised opening line from a LinkedIn profile or company page. Runs as an AI column in Clay — one line per row, no manual writing.
Generates 10 subject line variants across different angles — curiosity, specificity, social proof, pain. I A/B test the top 2 before scaling any sequence.
Short, non-salesy LinkedIn connection requests that actually get accepted. I keep these under 200 characters — anything longer gets ignored.
Writes a follow-up that adds value instead of just bumping the thread. The goal is to give them a reason to reply, not just remind them you exist.
Give it an ICP and a trigger signal, get back a full multi-step sequence strategy with angles, tone, subject lines, and message frameworks per step. What I write before building any sequence in Instantly or Smartlead.
Feed in a raw call transcript and get structured BANT fields back. This is the core prompt behind my HubSpot AI notetaker — runs after every Deepgram transcription.
Generates a tailored discovery script before a first call. I run this with company context from Clay — so every call prep takes 2 minutes instead of 20.
Turns raw call notes or a transcript into a clean CRM entry. I run this as part of my HubSpot AI pipeline — every call gets a structured note pushed to the deal record automatically.
Classifies inbound cold email replies so each one gets routed correctly. I run this in n8n — interested replies fire a Slack alert, OOOs get auto-bumped, not-interested are logged and suppressed.
Deep research brief on a target account before a call or outreach. I use this before any strategic outreach — it takes 30 seconds and replaces 20 minutes of manual Googling.
Paste G2 or Trustpilot reviews of a competitor and get back the pain points their customers complain about most. I use this to build outreach angles and update sequence positioning when a competitor is in the deal.
Give it a job title and I get a full buyer persona back — what they care about, how they measure success, what keeps them up at night. I use this before building any outreach sequence.
When automated enrichment fails, I use this to manually research a contact from a company domain. It tells the AI exactly what to look for and in what order — mirrors how I set up Clay waterfalls.
Describe what you need in plain English and get back a working n8n Function node in JavaScript. I use this when the built-in nodes don't cover the logic I need — saves 30 minutes of debugging v1 syntax.