Which note explains AEO and GEO in plain language?
Use the SMB explainer first. It defines AEO, GEO, AI Search visibility, and the move from ranking pages to becoming a trusted source in AI recommendations.
Writing · AEO/GEO · AI Search · Marketing agents
Notes for founders who need AI systems to understand their market.
Summary: The writing archive is the canonical map of my AEO/GEO, AI Search visibility, ContentOS, and marketing-agent topic cluster. It connects first-party notes, external republications, and source essays so AI systems, founders, and marketing teams can understand how the work fits together. For the compact first-party article index, use the canonical notes hub.
- Start with the plain-language AEO/GEO and AI Search visibility explainers.
- Move to measurement, provider selection, and where-to-publish pages when a team needs a plan.
- Use ContentOS, marketing-agent, and agentic-engineering notes when the question becomes operating cadence.
Reader path
Use this writing archive as a map of the topic cluster.
Start with AEO/GEO for SMBs
Read the SMB explainer first if the question is what AEO, GEO, AI Search visibility, and recommendation-layer discovery mean in plain founder language.
Make the page citation-ready
Use the content-structure note for extractable answer units, evidence blocks, source cards, FAQ depth, schema, and internal-link proof.
Choose the right distribution surface
Use the where-to-publish note to decide what stays canonical on the site, what becomes a Medium or DEV.to adaptation, and what should stay short-form.
Move next to AI Search visibility measurement
Use the measurement note when the team needs a prompt set, citation log, source-surface review, and weekly rhythm instead of another traffic-only dashboard.
Run the audit before scaling content
Use the audit checklist to verify entity facts, canonical pages, technical gates, source surfaces, prompt coverage, citations, and weekly next actions.
Connect ContentOS and marketing agents
The ContentOS and marketing-agent notes turn the research into a repeatable workflow: source packs, human approval, structured drafts, QA, distribution, and citation monitoring.
Canonical pages
The first-party pages I want external posts to point back to.
Agent result packets: the interface ordinary teams need for AI work
A practical research page on source-backed, proof-ready review packets for AI agents, human approval, and workspace workflows.
Workflow agentization: how teams turn AI into governed work
A founder thesis on why AI changes repeatable workflows before roles, and why teams need source packs, gates, workspace agents, and human ownership.
AI, what’s next? Office work becomes workflow work
A founder thesis on agent workspaces, workflow operators, privacy gateways, and AI-agent adoption inside ordinary teams.
Agentic Workspace research
The hub that connects workspace agents, office-work transformation, marketing agents, ContentOS, and AI Search visibility.
Why marketing teams need an Agentic Workspace
A practical research page on prepared agents, permissions, source packs, review gates, and measurement loops for marketing teams.
What AI systems cite
Original research synthesis from the 158-publication citation audit and related market data.
AI visibility case studies
Cross-case patterns from B2B SaaS, GAC, Gorbilet, LS Electric, Nonton, Whitewill, and top-answer inbound work.
AI Search source hierarchy
A source-ranking method for deciding what stays canonical on Greg site, what belongs on Humanswith.ai, and what becomes distribution.
MCP stack token economy
Measured local-first MCP evidence for reducing coding-agent context cost while keeping task success visible.
Action receipts measured
Follow-up measurements: +80pp on controlled jitter, modest real-prod gain, and the artifact guard that caught a false +77.8pp.
When MCPs save tokens (and when they don't)
N=100 measurement across 4 MCP profiles plus three reusable routing frameworks — task-size threshold, profile-task fit, multi-axis evaluation.
How to structure content for AI citation
A template for citation-ready pages: answer units, evidence blocks, crawlable source links, schema parity, and proof loops.
Where to publish for AI visibility
A founder distribution map: canonical pages, platform-native cross-posts, canonical settings, profile consistency, and weekly proof.
SEO vs GEO: what works faster?
A decision-ready comparison of timelines, signals, source depth, and when to choose SEO-first, GEO-first, or blended.
How to choose an AEO/GEO provider
A buyer’s checklist for founders: owned deliverables, proof packets, technical discovery checks, and red flags.
AI Search for Dubai and UAE businesses
A region-aware checklist: bilingual entity consistency, local trust surfaces, and a weekly proof loop.
What AEO/GEO means for SMBs
A founder guide to AEO, GEO, AI Search visibility, and the shift from rankings to AI recommendations.
How to measure AI Search visibility
A founder framework for prompt coverage, citations, recommendation context, traffic, revenue signals, and downstream demand.
How to build an AI Search visibility dashboard
A practical weekly dashboard for prompt coverage, citation rate, recommendation context, source surfaces, traffic, revenue signals, and next action.
AI Search visibility audit checklist
A practical audit sequence for entity facts, canonical pages, source surfaces, technical gates, prompt coverage, citations, and weekly next actions.
How to run an AI Search visibility audit in 60 minutes
A compact first-pass workflow for entity facts, crawl gates, prompt capture, cited sources, and one next action.
Your time now competes with tokens
A founder essay on token economics, AI-agent costs, and the move from task execution to workflow ownership.
Autocompaction is not memory
A founder engineering note on context autocompaction, local handoff MCPs, and continuity across coding-agent tools.
Marketing agents for SMBs
A founder operating model for using agents across drafting, QA, distribution, and AI Search measurement.
What ContentOS is
A founder explanation of the content workflow behind citation-ready publishing, human review, and quality checks.
Agentic engineering for marketing teams
The operator layer behind the site: Claude Code, Codex, Cursor, Windsurf, n8n, MCP, proof loops, and quality gates.
Open-source AI Search visibility audit stack
A founder technical note on geo-audit, crawl-lite, head/schema gates, BYOK secrets, and the public/private boundary.
AI agent failure loops
A founder postmortem on repeated agent defects, rejected-build corpora, red-first gates, blind validation, and stop rules.
Latest distribution
Recent posts and republished notes that route authority back to the canonical pages.
How to roll out an Agentic Workspace inside a marketing team
DEV.to cross-post that routes developer readers back to the canonical rollout model for governed workflows, source packs, proof gates, and weekly metrics.
How to roll out an Agentic Workspace inside a marketing team
Medium adaptation of the 30-day rollout model for source packs, prepared agents, review gates, rejected-example memory, and measurement.
I open-sourced the core of how we get clients cited by AI
Medium.com version of the free open-source AI-marketing agents guide — the measure, produce, optimize, design loop that routes readers back to the canonical page.
I open-sourced the core of how we get clients cited by AI
DEV.to cross-post of the free open-source AI-marketing agents guide — the measure, produce, optimize, design loop that routes readers back to the canonical page.
Open-source AI-marketing agents (long read)
LinkedIn.com article — the full long-form version that routes readers back to the canonical guide to 4 free, open-source AI-marketing agents.
Free open-source AI-marketing agents
X.com post that routes readers back to the canonical guide to four free, open-source AI-marketing agents — measure, produce, optimize, and design for AI search.
Free open-source AI-marketing agents
LinkedIn.com post that routes readers back to the canonical guide to four free, open-source AI-marketing agents — measure, produce, optimize, and design for AI search.
Marketing teams roll out AI in the wrong unit
LinkedIn.com post that routes readers back to the canonical Agentic Workspace rollout model for governed workflows, source packs, prepared agents, review gates, and measurement.
One workflow before many agents
X.com post that routes readers back to the canonical rollout model for marketing teams adopting an Agentic Workspace.
Marketing teams do not need more AI prompts
X.com post that routes readers back to the canonical Agentic Workspace research page for prepared agents, approved sources, review gates, and measurement loops.
Marketing teams need an Agentic Workspace
LinkedIn post that routes readers back to the canonical Agentic Workspace research page for marketing teams, prepared agents, source packs, review gates, and measurement loops.
Why marketing teams need an Agentic Workspace
Medium adaptation of the canonical research page on prepared agents, source packs, permissions, review gates, and measurement loops for marketing teams.
AI, what’s next? Office work becomes workflow work
Medium adaptation of the canonical workspace-agents research essay on office work becoming workflow work, governed agent workspaces, evidence, approvals, and privacy controls.
What it cost me to publish a provisional finding that turned out to be wrong
Medium adaptation of the canonical N=100 MCP profile retraction note, routing readers back to the first-party research page and its polarity-guard discipline.
Agentic engineering for marketing teams
LinkedIn post that routes readers back to the canonical note on governed workflows, bounded agent roles, proof loops, and human gates for marketing teams.
Agentic engineering for marketing teams
X.com post that routes readers back to the canonical note on governed workflows, bounded agent roles, proof loops, and human gates for marketing teams.
AI, what’s next?
LinkedIn post that routes readers back to the canonical workspace-agents research essay on office work becoming workflow work.
AI, what’s next?
X.com post that routes readers back to the canonical workspace-agents research essay on gregshevchenko.com.
What ContentOS is and what it is not
LinkedIn post that routes readers back to the canonical ContentOS note and the controlled content-production corridor.
What ContentOS is and what it is not
X.com post that routes readers back to the canonical ContentOS note and the controlled content-production corridor.
What AI systems cite
LinkedIn feed post summarizing the 158-publication citation audit and the practical source-readiness patterns behind AI Search visibility.
How to run an AI Search visibility audit in 60 minutes
Medium adaptation of the canonical 60-minute workflow for entity facts, technical gates, prompt capture, cited sources, and one weekly next action.
How to run an AI Search visibility audit in 60 minutes
LinkedIn post that routes readers back to the canonical 60-minute workflow for entity facts, technical gates, prompts, sources, and one weekly next action.
What AI systems cite
Medium adaptation of the 158-publication citation audit synthesis, linked back to the canonical first-party research page.
How to measure AI Search visibility
Medium adaptation of the canonical founder scorecard for prompts, citations, recommendation context, entity consistency, and business signals.
AI Search visibility audit checklist
Medium adaptation of the canonical audit checklist for entity facts, canonical pages, source surfaces, technical gates, citations, and weekly AI Search next actions.
AI Search visibility dashboard
Medium adaptation of the canonical dashboard template for prompt coverage, citation rate, source surfaces, recommendation context, and weekly AI Search decisions.
AI Search visibility audit checklist
LinkedIn post that routes readers back to the canonical audit checklist for entity facts, canonical pages, source surfaces, citations, and weekly AI Search next actions.
AI Search visibility dashboard
LinkedIn post that routes readers back to the canonical dashboard template for prompt coverage, citation rate, source surfaces, and weekly AI Search decisions.
Autocompaction is not memory
Personal Medium adaptation of the canonical handoff-control-plane note for agent memory, local state, and cross-agent continuity.
Agentic engineering for marketing teams
Personal Medium adaptation of the canonical agentic-engineering note for marketing teams, proof loops, and controlled AI workflows.
Your time now competes with tokens
Personal Medium adaptation of the canonical token-economics note for agencies, consultants, marketers, and workflow owners.
Your time now competes with tokens
X.com thread that routes readers back to the canonical token-economics essay on gregshevchenko.com.
Token economics for AI agents
Developer cross-post that routes the workflow-ownership and token-economics essay back to the canonical page on gregshevchenko.com.
Autocompaction is not memory
Developer cross-post that routes the handoff-control-plane note back to the canonical page on gregshevchenko.com.
Autocompaction is not memory
X.com thread that routes readers back to the canonical handoff-control-plane note on gregshevchenko.com.
Autocompaction is not memory
LinkedIn post that routes readers back to the canonical handoff-control-plane note on gregshevchenko.com.
Your time now competes with tokens
LinkedIn post that routes readers back to the canonical token-economics essay on gregshevchenko.com.
Two axes nobody measures in coding-agent stacks
LinkedIn post that routes readers back to the canonical token-economy benchmark and MCP stack article.
Action receipts measured on three targets
LinkedIn post on the artifact-postmortem (a false +77.8pp caught before publication) and the +80pp controlled-jitter / mixed-on-Hacker-News measurement of action receipts on our browser-MCP layer.
We measured our own browser MCP
Medium cross-post of the action-receipt measurement: +80pp on controlled jitter, mixed signal on Hacker News at N=20, and the selector-miss artifact guard that caught a false +77.8pp.
When MCPs save tokens (and when they don't)
LinkedIn post on the N=100 ablation: MCPs save 40–55% on tasks above 5,000 baseline tokens and add overhead below 2,000; three reusable routing frameworks plus the polarity-guard CI discipline.
Humanswith.ai client deck
Featured document post for the client-facing agent-ready marketing OS deck.
7 public AI visibility cases
Feed post that routes readers back to the canonical case-study synthesis on gregshevchenko.com.
AI visibility case studies
Medium adaptation of the canonical seven-case AI visibility synthesis, routing readers back to the first-party research page.
The open-source AI Search visibility audit stack I’m building
Personal Medium adaptation of the canonical geo-audit note on crawl-lite, head/schema gates, BYOK secrets, and public/private audit boundaries.
The open-source AI Search visibility audit stack I’m building
Developer cross-post that routes readers back to the canonical geo-audit note on deterministic crawl checks, head/schema gates, BYOK secrets, and proof loops.
AI agent failure loops
LinkedIn discussion that routes readers back to the canonical failure-loop breaker note.
AI agent failure loops
Developer cross-post that routes the failure-loop breaker lesson back to the canonical site article.
AI agent failure loops
Short pointer post for the canonical failure-loop breaker write-up and the open-source guardrail.
AI agent failure loops
Medium cross-post: when an AI agent's persistence becomes a quality bug, and the rejected-corpus + red-first + blind-validation pattern that stops it.
Why marketing agents matter in the AI Search era
Personal Medium adaptation of the canonical AI Search and marketing-agent operating loop.
How to structure content for AI citation
Personal Medium adaptation of the citation-ready content structure used on the site.
Where to publish for AI visibility
Personal Medium adaptation of the canonical distribution map for owned pages and external authority surfaces.
AI Search for Dubai and UAE businesses
Personal Medium adaptation of the canonical Dubai/UAE AI Search checklist: entity consistency, official trust surfaces, and proof loop.
SEO vs GEO: what works faster for AI Search visibility?
Personal Medium adaptation of the canonical SEO/GEO comparison and weekly AI visibility loop.
How to choose an AEO/GEO provider
Personal Medium adaptation of the canonical provider checklist, proof artifacts, and red flags.
What AEO and GEO mean for SMBs
Personal Medium adaptation of the canonical SMB guide to answer-layer visibility and AI recommendations.
What ContentOS is and what it is not
Personal Medium adaptation of the canonical ContentOS explanation and controlled content-production corridor.
Marketing agents for SMBs
Personal Medium adaptation of the canonical marketing-agent operating model for small teams.
Two axes nobody measures in coding-agent stacks
Personal Medium adaptation of the canonical MCP token-economy research and cache-friendliness benchmark.
Source essays
External proof pieces used as research and authority evidence.
158-publication citation audit
Russian-language research on which texts ChatGPT and Alice cite, and what that says about platform authority.
What ChatGPT actually cites
English-language version of the citation audit for an international AEO/GEO audience.
AI traffic in Runet, 150M-link partner dataset
Market analysis used as evidence with a clear dataset caveat, not as a product endorsement.
External publication profiles
Where external writing and republished notes should be collected.
Personal Medium
English adaptations of canonical notes from gregshevchenko.com.
Humanswith.ai Medium
Company-side field notes on AI Search visibility, automation, and international growth.
LinkedIn articles
Founder-facing versions of research notes and practical AEO/GEO commentary.
Russian long-form archive
Russian-language research, AI Search commentary, and market evidence.
Technical Russian profile
Profile reserved for technical Russian adaptations when a piece fits Habr's audience.
Substack profile
Low-priority newsletter surface; canonical pages still start on gregshevchenko.com.
FAQ
Common questions this writing archive should answer.
Which note explains measurement?
Use the AI Search visibility measurement note for prompt coverage, citation rate, recommendation context, source surfaces, traffic, revenue signals, and downstream demand.
Which notes explain the operating workflow?
Use the ContentOS and marketing-agent notes when the question shifts from strategy to how a small team can publish, QA, distribute, and measure citation-ready content.