A public operating system for how I work — Metrics, Analysis, Action. How I use AI, how to work with me, and how you can write your own portable context so you can move conversations across Claude, ChatGPT, Gemini, and any agent.

I use AI agents live to write articles, publish blog posts, build knowledge bases, and run SEO audits. I share my 4-stage Content Factory, the Task Library, and the magic phrase "according to" that makes AI follow my process.

Most marketers are still treating Claude like a fancy Google search. Here’s what you’re missing.

I’ve been in digital marketing for over a decade. I have witnessed the rise and fall of various tools, and I have honed a keen ability to distinguish between those that truly make a difference and those that merely sound impressive in a LinkedIn post.

Claude Code is different. Not because of the hype around it, but because of how it actually changes the way I work day to day.

When I first started using it, I was doing what most marketers do: typing prompts, getting outputs, and copy-pasting them into documents. That’s using a Ferrari to go grocery shopping.

The real shift happened when I started building a system around it. Now I don’t just use Claude. It runs alongside my entire marketing operation.

Here’s exactly how that works.

...

Everything that caught my attention in AI × Marketing last week ✨ LinkedIn becoming the #1 AEO channel, Anthropic's growth marketing team of one, HubSpot's personalization report, and the question I can't stop asking: are we building for agents or humans?

...

I run a small B2B agency and was trying to automate most of my work, writing ad copy, creating social content, get insights from performance data faster

so three months ago I decided to test every AI marketing tool that promised to "save time" or "automate" something meaningful

I spent $1,847 and gave each one a real 4-week trial on active campaigns

The pitch is always the same: AI writes your copy, designs your graphics, analyzes your data, generates insights- you just review and publish

that's not how it actually works, and I'm gonna be specific about why most of these tools are time-sinks pretending to be time-savers

Profound ($600/month): I tested it because my CMO saw a demo and it looked incredible. The dashboard is genuinely beautiful. I ran an analysis of our top-performing campaigns and it spit out attribution models that looked scientific. Then I manually checked the numbers and they didn't match our actual conversion data. Spent 8 hours trying to understand their methodology before support went silent when I asked direct questions. Killed after week 2.

Canva Magic Studio ($13/month): This one actually worked, but not how I expected. I thought I'd describe a campaign and it would auto-generate everything. In reallity it's a much better design tool than Canva was before, with some smart templates. But I still had to brief it properly, review every output, and fix copy. Time saved: maybe 20 minutes per week if I'm generous. Still paying for it because the design quality is legit, but it didn't change my life tbh

HubSpot's AI Features (included): The subject line generator works okay for email. The content assistant is surface-level. If you're already paying for HubSpot, sure, click the AI button- but it's not a reason to use HubSpot

Notion AI ($10/month): This one surprised me. I actually use it every day for things that aren't "AI magic." I use it as a CRM, a content calendar, and yeah, sometimes the AI fills in database fields or generates first drafts. Never once saved me hours. But the system itself (Notion, not the AI) reduced context-switching because everything lived in one place

Zapier (free tier): This is the one that actually moved the needle for me. It connected my existing tools so I wasn't manually copying data between systems. One workflow: new lead in my form, auto-filled contact in Notion, auto-triggered email sequence. Setup took 90 minutes and saves maybe 5 hours per month, pretty good!

Ryze AI ($49/month): They promise "AI that watches your ad campaigns and gives advice." What you get: alerts when performance drops, and a chatbot that gives obvious advice. Is your CTR down? "Try improving your ad copy or targeting." Unsubscribed after the trial

AI tools save time at the margins, not the fundamentals

they make a small job slightly faster. They don't eliminate 4 hours of work

the real time-saver was hiring a part-time person to do data entry and basic copywriting ($1,200/month)

that moved the needle way more than all six tools combined. But that's the honest conversation nobody has because there's no commission on recommending hiring someone

Every founder and marketer needs to set up AI workflows in 2026.

Not optional anymore.

The people who figure this out first are going to have an absurd advantage over everyone else.

I recorded a full walkthrough of the setup I use to run most of my marketing.

Ads, SEO, emails, community, thumbnails, dashboards. I'll show you how all of it works and how you can set it up yourself.

Four years of using AI every day = four years of pattern recognition across LLMs, four years of observing every bad habit AI creates.

We are at the stage where everyone is using the same descriptions for AI's impact on us.

Taste.
Judgement.
Agentic vs. Human

If you believe the nonsense X throws at you, you'll become the underclass if you don't learn to build agents.

I'm going to share some thoughts as an actual AI Native user. What does that mean?

I do most of my work using a terminal versus a prompt window. I program documents using MCP plugins. I automate everything I can - mainly low value work and I try to reduce context switching by limiting the number of tools I use by embedding as much as possible into Claude.

Let's start with judgement.

Here's a sample brief (I made it up). You're asked to create a brand for a new skincare range that ignores all the slop on Instagram. No lying, no deepfakes, no inaccessible beauty standards. Just a high quality product sold honestly.

As a CMO, you can hire an agency and the best work is still done by agencies IMO. A good strategist is worth the cost. The best brands in the world sit in your subconscious. That output does not come from an LLM. It's a human either prompting an LLM or using good old creative thinking. Anthropic use an agency and developers for that matter.

But if you're in a start-up and need a positioning statement or a communications platform, there are things you can do with an LLM to produce this thinking in-house. This is one of many tricks I have as a CMO trained in brand strategy. I cut a few corners here but this is a great example of "Judgement." AI speeds up the process. It does not produce this for you out of the box. Most people prompt "write me an opening line" and pick one. That is not how marketing works. AI is nowhere near replacing marketing. Ignore Anthropic when they tell you they have marketing skills - they do not. I would refer to them as task management tools.

So what is this doc? (Zoom in.)

These are the top 20 creative writing frameworks. If you have a clear product brief and you know your target customer, you can use these with AI to develop your own positioning statement and potentially a communication framework. I consider these outputs a V1 in terms of quality.

I named the brand STATED. Why? It states what it does instead of hiding behind AI slop advertising.

Despite what people say - education matters. Experience matters. You can't skip that phase of your career or the work - but you can learn some new ways to work faster when you need to. That is how judgement is developed - obvious statement but I think the internet forgot.

Save these frameworks for later - they might come in handy ✍🏻

I test a new AI tool every week, and this month 10 earned a permanent spot in my workflow. I used to spend hours doing things manually until I started systematically experimenting. The tools I kept help me move faster across writing (voice-to-text for long emails, WhatsApp replies, and high-context prompts), meetings (invisible AI note-taking for sensitive client calls), research (source-grounded summarization and internal-doc audio overviews), automations (agent/workflow builders for recurring research and backend workflows), and content production (transcript-based video editing, AI voice generation/repairs for marketing voiceovers, and video generation models for cinematic clips).