← Field notesAgentic

Running a one-person marketing department with AI

Ryan Walker8 min readUpdated June 2, 2026

Running a one-person marketing department with AI

Six marketing functions keep a business growing: content, SEO/GEO, paid media, email, social, and analytics. In 2026, each one has a working AI solution. The question is not whether to use AI for marketing — it is how to wire the six functions together so they run without you.

Content: the 80/20 workflow

AI drafts; you edit. That ratio is the whole model.

The workflow: write a 150-word brief — topic, angle, target reader, one key claim. Feed it to your AI writer. Get a full draft back in under two minutes. Spend 20 minutes editing for accuracy and voice. Run a critic pass (ask the model to find weak claims and unsupported assertions). Publish with GEO structure: answer-first opening, FAQ block, named author, freshness stamp.

Time per post: 30 minutes. Daily publishing is achievable for a single operator. Most one-person teams are not publishing daily because they are still writing from scratch.

SEO and GEO: the weekly audit agent

An AI agent audits your top 10 pages every week. It checks five things: answer-first structure, freshness stamp, named author, FAQ block, and citation signals (clean factual claims an LLM can lift verbatim). It flags what needs updating and ranks the flags by estimated impact.

You spend 30 minutes per week acting on the flags. You do not read the pages yourself. You do not run the audit yourself. The agent surfaces the delta; you make the call.

This is the difference between SEO as a project and SEO as a system.

Paid media: agentic PPC

Google Ads Performance Max, Meta Advantage+, and dedicated agent layers now monitor bids, search terms, and creative performance in real time. They adjust without you.

Your job is three inputs: budget, target CPA, creative brief. The agent handles bid strategy, audience expansion, creative rotation, and negative keyword management. You review a weekly summary — 15 minutes — and update the creative brief when performance plateaus.

The failure mode is treating agentic PPC like manual PPC and over-riding the automation. Set the guardrails, then let it run.

Email: sequences from briefs

AI writes email sequences from a brief. A brief is 200 words: audience, goal, tone, number of emails, trigger event. The model returns a full sequence — subject lines, body copy, send timing.

You approve the first send. The rest run on triggers: welcome sequence fires on signup, nurture sequence fires on content download, re-engagement sequence fires on 60-day inactivity.

Time investment: one brief per sequence, one approval. Ongoing time: zero. The sequences run until you change them.

Social: caption batching

Every blog post generates three social captions via a caption agent — one for LinkedIn, one for X, one short-form option. You pick one, edit if needed, and schedule it in Buffer or Hootsuite.

Time per week: 20 minutes, assuming four posts per week. The captions are not an afterthought; they are a byproduct of the content workflow. No separate social strategy required.

Analytics: the weekly memo

An AI agent reads your GA4 data every week and produces a plain-English memo: what changed, why it changed (correlated with traffic sources, content, and campaigns), and what to do about it.

You read it in 10 minutes and make three decisions. Which page to update. Which channel to push. Which experiment to run next.

The agent does the reading; you do the deciding. That division of labor is the point. Most analytics time is spent pulling data, not acting on it. The agent eliminates the pull.

We send the exact prompts and agent configs for each of these six functions to Field Notes subscribers. Get them at avakata.agency/contact.html.

Total time investment

Here is what the full operation costs per week:

  • Content: 30 min/day (3.5 hours/week at five days)
  • SEO/GEO audit: 30 min/week
  • Paid media: 30 min/week
  • Email: 2 hours/quarter (negligible weekly)
  • Social: 20 min/week
  • Analytics: 10 min/week

Total: roughly 4 hours and 50 minutes per week for a full marketing operation across six functions. That is not a part-time job. It is a Tuesday morning.

The constraint is no longer headcount. It is whether you have the systems configured. If you want to audit your current setup against this model, book a discovery call and we will walk through it.

Frequently asked questions

Can a solopreneur run a full marketing operation with AI?
Yes. Content, SEO/GEO, paid media, email, social, and analytics each have working AI solutions in 2026. A solopreneur with the right stack spends roughly four hours per week governing a marketing operation that would previously have required a team of six. The human handles judgment and decisions; the AI handles volume and execution.
How much time does AI marketing take per week?
Roughly four hours per week for a full marketing operation: 30 minutes per day on content editing, 30 minutes per week on SEO/GEO audit actions, 30 minutes per week on paid media review, 20 minutes per week on social scheduling, and 10 minutes per week on analytics review. Email sequences are set up quarterly and run on triggers.
What AI tools do solopreneurs use for marketing?
Content: Claude or ChatGPT with a brand voice prompt. SEO/GEO: a custom audit agent or tools like Avakata. Paid media: Google Performance Max, Meta Advantage+, or an agentic PPC layer. Email: Klaviyo or ActiveCampaign with AI sequence generation. Social: Buffer or Hootsuite with a caption agent. Analytics: a GA4 summary agent or tools like Polymer.

Related reading