We’re all using ChatGPT, Gemini, Claude… but let’s be honest — most of what’s coming out of these tools is rubbish. Generic, unfunny, over-formal slurry. M-dashes everywhere. Worse still, it’s killing brands and cluttering our inboxes with AI-written drivel that nobody reads.
So what if you could build an agent that works like a team member? One that understands your tone, writes like you, and plugs straight into your content engine? That’s exactly what Reed Hansen helps us uncover in this episode. Reed’s a Chief Growth Officer and strategic marketing brain who’s helped everyone from Oracle to fast-growing startups. He breaks down exactly how to build and train an AI agent that’s not only useful — but actually saves you time, scales your voice, and makes you look smarter.
We talk about how most people are doing this all wrong, the real value of AI agents, and how to avoid turning your brand into a slurry cannon. This episode is crammed with practical advice you can use right now — especially if you're trying to do more with less and keep your content engine running without sounding like a robot.
How to build your own AI marketing agent (the right way)
✅ Define the agent’s role and constraints: Who’s it for? What outcome do you want? Be specific about tone, topics to avoid, and the format of the content.
✅ Train it like a new hire: Don’t just say “write like me” — give it FAQs, onboarding docs, examples, brand voice guidelines, and preferred content structures.
✅ Attach high-trust sources: Feed it with your own high-quality material — case studies, blog posts, sales copy, anything good. Don’t let it guess.
✅ Connect it to your workflow: Use tools like Zapier or Make to push content into Buffer or your CRM. If you're stuck copy-pasting, you're not there yet.
✅ Supplement with your face: AI can't be you. Mix in your voice, your face, your real thoughts. That’s what builds trust and gets remembered.
Timeline summary
[01:54] – Why most teams fail: it’s not the reps, it’s the ICP or the manager [02:44] – Agentic AI is “the dog’s proverbial” — why it’s the next step beyond ChatGPT [04:53] – Treat it like an employee: how to shape responses with consistent training [07:44] – The M-dash and the curse of legalese: where bad AI writing comes from [08:44] – What people get wrong: not enough instruction, too much freedom [10:24] – Chat is not enough: agents need to push/pull data to be truly valuable [11:08] – Three-step framework: plan, build the engine, and connect it [14:21] – Structuring your prompts: listicles, tables, tone, formats — get specific [15:31] – Plug into your planner: automate workflows to get daily content out [20:13] – The AI echo chamber: why we’re amplifying internet sewage [24:46] – Use AI for hygiene content, but mix in real human posts too [26:02] – “Put your face on the content” — realness beats slurry every time [28:21] – Final advice: your face + your voice = your brand’s secret weapon
Links & resources
The Practical Leadership Academy – Free Parallel Team Playbook
Buffer – Social media scheduling tool
Zapier – Automation between apps
Make – Visual automation builder
If you found this episode valuable, please do me a favour — rate, follow and review the podcast. It really helps us get the word out to more sales leaders who need to hear it. And don’t forget to share it with someone who’s still letting AI write like a UN lawyer with a thesaurus.
Show more...