
Advanced Guide to Building CustomGPTs For ChatGPT
Build a focused CustomGPT around a repeated marketing job by defining when to use it, what it should do, what it should ask, and which reference material sets the standard.
Open episode →AI-Driven Marketer podcast
Browse the complete archive of practical conversations and experiments for marketers building with AI.
Episode archive
Page 13 of 18

Build a focused CustomGPT around a repeated marketing job by defining when to use it, what it should do, what it should ask, and which reference material sets the standard.
Open episode →
A useful AI stack is smaller than a tool directory. Commit to a general system, keep tools that solve real bottlenecks, and let repeated results decide what stays.
Open episode →
Use AI to shape stories grounded in your real experience, connect them to the customer journey, and adapt them across marketing without flattening your voice or losing the human reason to care.
Open episode →
The o1 preview model adds deliberate reasoning to ChatGPT. Use it for difficult marketing decisions, supply context, and keep human judgment in the loop when reviewing its recommendations.
Open episode →
A practical framework for using AI at the constraint that matters most, from faster execution and better analysis to focused market choices and new possibilities.
Open episode →
The most useful AI practice starts with ordinary problems. Explore how niche research, custom assistants, screenshots, images, and experiments build practical judgment.
Open episode →
AI news can create hype and FOMO. I share the 2024 sources I used for curation and context, then explain why hands-on experiments matter more than simply following the headlines.
Open episode →
CustomGPTs and niche-trained language models can extend a marketer's expertise, but neither replaces the judgment required to diagnose problems, ask questions, and make decisions.
Open episode →
Craig Nelson and I discuss a practical roadmap for AI-driven sales teams, from call preparation and follow-up to proving time savings and institutionalizing repeatable workflows.
Open episode →
I tested Grok2's access to live X conversations for market research, from finding B2B trends and complaints to shaping content ideas and checking sentiment in a faster research loop.
Open episode →
AI gets more reliable when you stop asking it to remember everything and instead load the right context, sources, instructions, and examples into its working memory.
Open episode →
SearchGPT signals a shift from ranking for clicks to being selected inside answers. Test the search behavior now and publish evidence AI systems and humans can trust.
Open episode →