How to Build an AI Agent (OpenClaw) That Runs Your GTM with Koka Sexton
B2B OG Koka Sexton Goes From Social-Selling Legend to AI Revenue Architect
This week is a two-for-one episode, which also makes it our longest one yet. Koka Sexton, an OG in B2B who helped pioneer the social selling movement, joins me for a great conversation. He’s leading the charge on agentic marketing and AI agents and walks through a few Make.com workflows in detail.
But what really grabbed me is what he’s building with OpenClaw. Full transparency: I haven’t installed it yet. I briefly considered it… then chickened out over the security questions. But honestly, it feels like it’s only a matter of time.
Anyway, here’s this week’s Stack:
• 1 deep dive: Build an AI agent to run your GTM with Make + OpenClaw
• 1 prompt: Turn AI into a structured “coworker” with guardrails and approval gates
• 1 tool: Detect whether content was written by AI
• 3 resources: AI ROI reality checks, strategic AI frameworks, and the future of agents
• 3 jobs: Product marketing, AI data, and AI-native GTM leadership roles
Let’s go.
Workflow Walkthrough: Build an AI Agent That Runs Your GTM
Initially, when I asked Koka to come on the show, he was going to show off some of his awesome Make workflows. But when it came time to scheduling, he said, “I have something even better to show you.” I said, “show me both!”
So in this episode, we dive deep, walking through his Make workflows and how he’s using OpenClaw.
Here’s what you’ll discover in this episode:
How to export LinkedIn likes, comments, and shares and turn them into high-intent pipeline and drop them into an automated Make workflow
When Zapier breaks and how 40-step Make.com automations take over
How to collapse 40K LinkedIn connections into a segmented, searchable CRM with OpenClaw
How OpenClaw routes tasks to the best model automatically
Why “bots talking to bots” is the next GTM frontier
Prompt of The Week: AI Coworker with Safety Rules (for Claude CoWork)
Most people (still) use AI like a chatbot. But I like to think of them as teammates. This is the kind of structured “coworker prompt” that the top markeres use. It has a clear task, clear inputs, strict safety rules, and explicit approval gates. You can use this same structure for outbound, content distribution, CRM updates, or internal reporting.
You are helping me [clear outcome].
In this folder you will find [describe structured inputs clearly].
Your job:
[Task 1]
[Task 2]
[Task 3]
For each item, you must:
[Personalization requirement]
[Formatting requirement]
[Output requirement]
After drafting everything, [execution instructions].
Important safety rules:
Only use [approved data source].
Do not [prohibited action].
Never execute without my explicit approval.
If anything is unclear, ask a clarification question instead of guessing.I’ve added this to my The Ultimate ChatGPT Prompt Library for B2B Marketing Leaders notion doc. Check it out for 60+ more prompts.
Tool of the Week: Is This Actually Written by a Human?
Almost everything I read now feels like it was written by AI. I’m talking about blog posts, LinkedIn posts, cold emails, website… you name it.
If you’re like me, you’re thinking to yourself, “AI wrote this, didn’t it?” Well, this tool gives you a quick answer. GPTZero analyzes text and estimates the likelihood it was generated by AI. Of course, it’s not perfect, and there are other tools that will be better, but this is a great free tool.
Try it: https://gptzero.me/
AI Resource Roundup
The Week in AI: ROI When Reality Bites Back (Ray Rike): A sharp breakdown of where AI hype meets financial reality. Critical if you’re defending AI spend to a CFO right now.
Almost Timely News: How I Think About AI (Christopher Penn): A thoughtful framework for evaluating AI tools strategically instead of chasing every shiny object.
The Next Wave of AI Agents (YouTube): A forward-looking discussion on agentic systems and where they’re headed. Helpful context if bots-talking-to-bots got your wheels turning.
Hot AI Jobs
Marketing Leader at DatologyAI
Location: Redwood City, CA (On-site)
Pay Range: $200K–$275K
Director, Product Marketing at Descript
Location: San Francisco, CA (Remote)
Pay Range: $220K–$250K
Marketing Director at Predactiv
City / Remote: Palo Alto, CA (Remote)
Pay Range: $125K–$170K
By now, I’m sure you’ve seen the posts about something big happening in AI. It definitely feels that way. If you’re like me, you probably oscillate between being really excited and overwhelmed.
I’m here to encourage you to go ahead. You don’t need to master everything. Just build one thing at a time.
Now, please excuse me while I go play around with Claude Code.
Brandon