Shaun Thresher built his own AI powered second brain, using Claude Code, to run his marketing and copywriting work end to end. In this episode, Shaun walks through how he moved from bouncing between two separate AI chat windows to a single centralised system that reads his voice, his clients' details and his workflows before it writes a word.
He explains why one well briefed agent with multiple skills beats a pile of specialised agents, how he trains AI to score new content against his best performing posts so it stops sounding repetitive, and how he uses AI to dig up prospect lists most competitors never find, including one built from a public FDA compliance database. Whether you're just starting to explore Claude or ChatGPT, or you're ready to build a proper system around your marketing operation, this conversation is full of practical, hard won lessons.
Guest Introduction
Shaun Thresher is a copywriter turned AI systems builder who spent eight years at Agora Financial, a top US direct-response shop, before becoming CMO of online assets at a publicly-traded company and later CEO of one of its e-commerce brands. He now runs BioRevGrowth, building AI powered "second brain" systems for marketers, and since January 2025 has led marketing and technical operations for a US-based long COVID doctor.
Key Topics
Moving from copying and pasting between ChatGPT and Claude to a single centralised second brain, built and stored in GitHub
The difference between agents, skills and workflows, and why Shaun now favours one well briefed content agent with multiple skills over many separate agents
How Shaun sources prospect lists from public databases, including an FDA compliance database, instead of relying solely on paid tools like Clay
Building a "digital fingerprint" from a client's best performing content so AI generated posts match their voice and tone
Scoring new content against a curated pool of top performing past content to avoid repetitive, generic AI output ("AI slop")
Why Claude Code offers more flexibility than a standard chat interface, including API and MCP connections into tools like Google Ads and Meta Ads
Common pitfalls: memory systems that quietly stop working, and installing unfamiliar agents without understanding how they operate
Advice for marketers starting out: pick one problem to solve, invest in a higher tier AI plan, and find a community for support
Podden och tillhörande omslagsbild på den här sidan tillhör
xGrowth. Innehållet i podden är skapat av xGrowth och inte av,
eller tillsammans med, Poddtoppen.
