The AI Marketing Department
We build and operate your AI marketing department on monthly sprints. £3,000 to £8,000 a month, no minimum. Each sprint, the system gets sharper as it learns your business. The work is not a project. We replace the marketing function — content production, paid acquisition, attribution, AI search, model selection — and run it under one operating layer on your infrastructure. Outputs ship every week, executed by agents in production, reviewed by humans, monitored against your real revenue numbers.
Implementation
How we set up the AI marketing department on your stack — accounts, agents, data flows, observability. The first sprint, end-to-end.
Sub-clusterInfrastructure
The platforms, integrations, and orchestration layer the function runs on. Always your stack — your cloud, your CRM, your Notion. We get access; you keep ownership.
Sub-clusterWorkflows
The actual production pipelines. Agent-led, human-reviewed, shipping weekly. From brief to draft to publish, from ad copy to placement to attribution.
Sub-clusterEconomics
What it costs to run the marketing function on AI versus hiring. Hours-based pricing, multi-model stack, no separate API invoice.
Sub-clusterFailure modes
The places AI marketing breaks — hallucinations, brand voice drift, attribution gaps, model deprecation. How we catch them, in production, every sprint.
Sub-clusterMost marketing teams are running stacks designed for a 2018 world — disconnected tools, manual workflows, and humans doing what AI can now do faster and cheaper. The marketing function is a department, not a tool problem. We rebuild it as one.
— Marketing Operations thesisMarketing Operations
Why most AI marketing automation projects stall at month 3
Why most AI marketing automation projects stall at month 3
Read post: Why most AI marketing automation projects stall at month 3Marketing Operations
The AI marketing operations playbook that's built for reality
The AI marketing operations playbook that's built for reality
Read post: The AI marketing operations playbook that's built for realityMarketing Operations
How we built our own rank tracker with DataForSEO + Claude Code
We built our own rank tracker with DataForSEO + Claude Code — £27/month versus £99+ for Ahrefs. Complete architecture, TypeScript implementation, PostgreSQL schema, cost breakdown, and the five things that broke during development.
Read post: How we built our own rank tracker with DataForSEO + Claude CodeContent Strategy
AI content production at scale, with quality and brand voice as constraints.
Explore Content StrategyAttribution
Server-side tracking, multi-touch attribution, and the models that survive iOS and GA4 changes.
Explore AttributionAI Search
Showing up where buyers ask AI — Google AI Overviews, ChatGPT, Perplexity, Claude — plus the AI-era SEO that still feeds them.
Explore AI SearchAI Models
Comparing Claude, ChatGPT, Gemini and others — and picking the right model for each marketing job we run for clients.
Explore AI ModelsBook a discovery call
Sixty minutes. No slide deck. We tell you which sprint tier fits and when output starts shipping.
Get in touch →