Why a Blog about Vibecoding for CMOs?

My LinkedIn feed is exhausting lately. Everyone’s got a post or ten about AI.

But when I talk to other CMOs, actually talk, here’s what I hear: board members want “AI transformation” (whatever that means), vendors fill my inbox with magical promises, and there’s this underlying anxiety about whether our experience as CMOs still matters.

I’m a CMO. I’ve marketed technical infrastructure to geeks for 20+ years. Always been the idea person who handed things off to builders. But I needed to know if I could still be useful in whatever this new thing is, so I started actually building stuff.

This blog is me documenting that. It’s called The Vibecoding CMO because “vibecoding” is what people are calling it when you describe what you want and AI writes the code. Honestly, the name might be lame but it stuck in my head.

The problem I keep running into

There’s tons of high-level AI content.

“Transform your marketing with AI!”

“Here’s how I replaced half of a 20-person marketing team with 45 AI agents (I actually saw this post on LinkedIn)

Okay, so the big question is how? Then there’s documentation that assumes you know what an API endpoint is. Also, API docs are notorious for being out of date. I do now, sort of, understand but that’s also kind of the problem. There’s this whole middle layer of knowledge that’s really hard to search for because you don’t know what you don’t know. I’ve spent more than half of my agent-building time trying to figure out how to get integrations to work and fighting with API keys.

Most of us marketing leaders are stuck in the middle. We need actual tactical guidance that doesn’t require a CS degree but goes deeper than “try ChatGPT!” The marketing tech stack is too important now to just hand decisions to IT and hope they get it. Things move too fast. And honestly, even when IT is great - and I’ve worked with some strong engineering teams, they have their own roadmap and priorities. Waiting three sprints for a simple marketing automation thing to get built feels insane, when you could maybe just do it yourself, if you only knew how. Except you don’t know how. Or you sort of know how but not really.

I’ve had versions of this conversation with enough other CMOs to know it’s not just me being dumb. Though sometimes it feels like it.

What I’m actually doing

Using AI coding agents (Claude, Cursor, stuff like that) to build marketing tools that supports my work as a CMO. What surprised me is that you still need to think like an architect. Not about writing code necessarily, but understanding how systems connect. Data flows. What’s actually possible versus what sounds good in a meeting.

We’re not becoming developers. We’re becoming... I don’t know, technical enough to be dangerous? Still figuring out what to call this.

Why you might care what I think

I’m a nerd at heart with an MIT MBA, along with chemistry and economics undergrad majors (random, I know). I was head of marketing at EDB, where we scaled from $25M to $100M ARR less than 5 years. I’m currently doing fractional CMO work with two companies: An open source infrastructure software company and an early stage medical imaging infrastructure startup.

The credentials aren’t the point though. What’s probably more useful is that I’ve sold to developers for years, worked closely with engineering teams, and now I’m trying to build alongside them.

What this blog will be

Real examples from what I’m building. Code when it’s relevant, skipped when it’s not.

Case studies from my current work (anonymized when needed). How to be technical enough for modern marketing leadership without pretending you’re an engineer.

I’m also testing something bigger—whether AI agents can actually execute GTM, not just help with it. Some of it will probably fail. You’ll see that too.

No hyperbole.

There’s already plenty of that on LinkedIn.

I should probably have a better call-to-action here, but honestly, I’m just going to start writing these and see if anyone reads them.