Writing.
Notes from building AI marketing systems, fractional CMO engagements and agentic GTM infrastructure in public.
Start here if you are exploring AI marketing systems.
The writing here documents the practical loop I am building: LinkedIn signal, autonomous article production, SEO/AEO improvement and commercial proof on the website.

Vercel vs Railway: A Practical Guide for Agentic Marketers
Vercel and Railway can both support websites, backend applications and AI agents. This guide explains their operating models, costs and practical marketing use cases.
AI marketing systems should start with signal
Why the useful first layer for AI marketing is continuous customer signal: the questions, frustrations and market changes that should shape content, campaigns and sales conversations.
AI marketing now needs architecture skills
Why senior marketers need to understand agent architecture, context, evaluation and approval design if they want AI marketing systems that affect commercial outcomes.
AI startups need data and distribution more than another agent
Why easier AI product building pushes defensibility towards proprietary data, positioning and distribution, and what founders should build around the product.
Loop engineering is how AI marketing becomes a system
Why loop engineering matters for AI marketing agents, and how founders can turn repeated marketing judgement into governed systems that improve through use.
Agentic marketing is becoming the fourth generation of marketing
Why agentic marketing changes the job of the senior marketer, and why founders should care more about systems, judgement and commercial outcomes than AI content volume.
llms.txt is useful infrastructure, not an AI discovery strategy
What this week’s LinkedIn posts showed me about llms.txt, AEO loops and the practical work required to make a brand easier for AI systems to find, trust and cite.
AI AgentsBad approval design is the real risk in AI marketing.
AI marketing systems fail when approval design, state, audit trails and publishing gates are too loose for real brand work.
AI AgentsI built a newsletter agent with an approval gate. Here is the operating system.
How I use Hermes, Tavily, MailerLite and a human approval gate to turn weekly AI marketing research into a newsletter and blog workflow.
AI AgentsI built an autonomous lead gen system with Claude, Hermes and Orange Slice.
How I turned Streamr's partnership and investor pipeline into a repeatable agentic GTM system with Claude, Hermes, Apollo, Clay, Orange Slice and GitHub.
AEOI tested whether AI engines could find me. They could not.
What I changed on mark-little.com after testing whether ChatGPT, Perplexity and Claude could identify me as a fractional CMO for AI marketing systems.
Get the weekly AI marketing briefing.
Agent research, human editing, and practical systems over AI news theatre.
- ✓Practical systems over AI news theatre
- ✓Short enough to read properly
No spam. Unsubscribe anytime.