Agent economy. Three domains. 504 keywords scored. 2,481 backlink targets. The window before incumbents arrive is open right now.
The agent economy has no SEO incumbents yet. The categories being invented now - agent skills, internet of agents, pheromone routing, bonding curve agents - have virtually no established content, no dominant domains, and zero LLM citations when queried. We invent and define these categories on our 3 domains today, so when LLMs ingest 2026 training data, our pages become the canonical source. This window is 12-18 months. Probably less.
Pheromone-routed signals. Sub-10ms latency at Cloudflare edge. 220-line minimalist schema. ONE SDK on npm. 40-tool MCP server for Claude + Cursor.
| Keyword | Score | KD | Vol US | Trend |
|---|---|---|---|---|
| agent payroll | 100 | 0 | inventing it | VIRGIN |
| nanoclaw | 90 | 18 | 2,400 | RISING |
| edge ai | 85 | 28 | 5,400 | RISING +82 |
| claude skills marketplace | 85 | 17 | 1,900 | SPIKE +248% |
| payment streams | 85 | 0 | 70 | RISING +157% |
| mcp server | 80 | 55 | 60,500 | SPIKE |
| claude skills | 80 | 50 | 27,100 | RISING +495% |
| tool router | 65 | 9 | 27,100 | RISING |
| agent identity | 80 | - | - | RISING +147% |
| agent treasury | 80 | - | - | RISING +157% |
Full 30 P1 + 79 P2 in scored-opportunities.csv. "agent payroll" score 100 = zero content exists anywhere. Pure category-creation bet. Click Cluster F above to see all.
Deploy an agent. Mint a token. Trade on-chain. Per-agent revenue, bonding curve economics, Agentverse integration.
| Keyword | Score | KD | Vol | Trend |
|---|---|---|---|---|
| pump fun for agents | 80 | - | inventing it | VIRGIN |
| virtuals alternative | 80 | - | inventing it | VIRGIN |
| deploy agent Agentverse | 80 | - | - | VIRGIN |
| bonding curve agent | 80 | - | - | VIRGIN |
| tokenise agent | 80 | - | - | VIRGIN |
| agent token | 55 | - | 20 | RISING +188% |
| agent coin | 55 | 33 | 50 | RISING |
| agent ICO | 55 | - | 10 | VIRGIN |
| agent fair launch | 55 | - | - | VIRGIN |
| tokenize AI agent | 55 | - | - | VIRGIN |
Volume is low because the category doesn't exist yet. That is the strategy. Pump.fun had no search volume before it launched either.
OpenAI GPT Store has almost no external backlinks - the marketplace model hasn't built link equity. Opportunity to leapfrog early.
SMB retainer $500-4k/mo. AI SEO, AEO audits, GEO optimization, local AI chatbot. Mid-market integrations at /integrations/ ($5-12k/mo). Cluster M at /internet-of-agents/.
12 original clusters A-L. Cluster M added April 2026 (Internet of Agents). Click a cluster card to expand its full P1+P2 keyword table.
504 keywords x 3 geos (US/UK/IE) x 5yr weekly interest scores. $2.74 total cost. 1509 kw-geo pairs classified. DataForSEO google_trends/explore/live.
US geo only. Total all geos: 5 SPIKE, 63 RISING, 1083 STEADY, 317 VIRGIN, 41 DECLINING across 1509 rows.
Sampled every 12 weeks from 262 weekly data points. Y-axis = Google Trends interest score 0-100.
| Keyword | Cluster | Growth |
|---|---|---|
| claude skills | C | +495% |
| machine commerce (SPIKE) | H | +440% |
| claude.md | I | +371% |
| claude code (SPIKE) | I | +305% |
| agent skills (SPIKE) | C | +248% |
| prompt caching | I | +240% |
| agent token | K | +188% |
| internet of agents | M | +159% |
| agent treasury | F | +157% |
| agent identity | G | +147% |
| geo optimization | J | +128% |
| agent commerce | H | +121% |
| Keyword | Cluster | Drop |
|---|---|---|
| agentkit | A | -75% |
| agentkit openai | A | -75% |
| x402 crypto | A | -70% |
| agentverse | B | -52% |
| best ai seo agency | OO | -48% |
| blockchain seo | E | -37% |
| agentic commerce companies | H | -35% |
| agent presale | K | -34% |
Note: "x402 crypto" declining but "x402 payment protocol" framing is different. Pivot the framing, not the technology.
Live SERP data for 25 P1 keywords + real domain authority for all competitors. Run 2026-04-30. Baseline: we rank for 0/25 keywords (domains brand new). That is where we start.
one.ie: 42 referring domains vs fetch.ai: 5,069. 120x gap. This is the backlink authority race we are running. Content strategy buys time while links are acquired.
| Keyword | Top 3 incumbents | Opportunity |
|---|---|---|
| pheromone routing | aps.org, arxiv.org, ieee.org | EASY - science papers |
| internet of agents | cisco.com, arxiv.org, github.com | MEDIUM - Cisco has a page |
| internet for agents | cisco.com, agntcy.org, medium.com | MEDIUM - agntcy.org competitor |
| GEO optimization | wikipedia.org(#6), reddit.com, reply.com | EASIEST WIN - no specialist |
| claude code vs cursor | reddit.com, codeaholicguy.com, ksred.com | EASY - pillar written, publish now |
| agent skills | agentskills.io, claude.com, github.com | MEDIUM - agentskills.io exists |
| AEO audit | aeoaudittool.com, aeoengine.ai | EASY - small early competitors |
| agent commerce | mckinsey.com, ibm.com, stripe.com | HARD - big brand authority |
| mcp server | modelcontextprotocol.io, reddit.com | MEDIUM - official Anthropic docs |
| edge ai | ibm.com, redhat.com, reddit.com | MEDIUM - ONE.ie angle beats IBM |
Links first observed in DataForSEO index by month (2024+). Shows competitor link-building acceleration. Not link creation date - DataForSEO index date.
50 category-relevant queries across ChatGPT (gpt-4o), Claude (claude-sonnet-4-6), Perplexity (sonar-pro), and Google AI (Gemma-3-27b-it via Google AI Studio API). Domain citations logged per response. Google AI note: Gemma-3-27b is Google's open-weight research model - not the same as Gemini-2.5-flash which powers Google AI Overviews. Gemini-2.5-flash was overloaded (503) on the day we ran. The Gemma data is valid citation research; re-run with Gemini-proper is queued. Q46-Q50 Perplexity re-run 2026-04-30 after credit top-up. 220+ total API calls.
Critical: all 6 of our citations are navigational (queries that literally named our domain). Category citations = 0. The race to own category answers hasn't started.
| LLM | Total citations | Our domains | Competitors |
|---|---|---|---|
| ChatGPT | 4 | 2 | 1 |
| Claude | 26 | 2 | 8 |
| Perplexity | 27 | 2 | 4 |
| Google AI (Gemma-3-27b) | 82+ | 0 | competitor hits (fetch.ai consistent) |
Q46 Perplexity: otterly.ai cited. Q47: seo.com cited. Q50: fastbots.ai cited (not ONE.ie - content gap).
Zero Google Trends data = either pre-discovery (no searches yet) or Google threshold not met. Both = first-mover advantage. Hover any tag for the content angle. Search to filter.
| Keyword | Cluster | Domain | Score | Content angle |
|---|---|---|---|---|
| outreach ai revenue agent | ? | ? | - | |
| agentkit download | A | ONE.ie | 55 | Definition-first explainer on 'agentkit download' with comparison matrix vs Coin |
| agentkit github | A | ONE.ie | 55 | Definition-first explainer on 'agentkit github' with comparison matrix vs Coinba |
| argent smart contract wallet | A | ONE.ie | 55 | Definition-first explainer on 'argent smart contract wallet' with comparison mat |
| argenthq | A | ONE.ie | 55 | Definition-first explainer on 'argenthq' with comparison matrix vs Coinbase, Cro |
| is agentkit free | A | ONE.ie | 55 | Definition-first explainer on 'is agentkit free' with comparison matrix vs Coinb |
| starknet wallet extension | A | ONE.ie | 55 | Definition-first explainer on 'starknet wallet extension' with comparison matrix |
| x402 scan | A | ONE.ie | 55 | Definition-first explainer on 'x402 scan' with comparison matrix vs Coinbase, Cr |
| deploy agent agentverse | AL | agent-launch.ai | 80 | agent-launch.ai product page for 'deploy agent Agentverse' with per-agent 120 FE |
| agentverse github | B | ONE.ie | 55 | Category-defining post on 'agentverse github' with vendor-neutral matrix, live O |
| agentverse google | B | ONE.ie | 55 | Category-defining post on 'agentverse google' with vendor-neutral matrix, live O |
| agentverse paper | B | ONE.ie | 55 | Category-defining post on 'agentverse paper' with vendor-neutral matrix, live ON |
| mcp marketplace smithery | B | ONE.ie | 55 | Category-defining post on 'mcp marketplace smithery' with vendor-neutral matrix, |
| anthropic skills jar | C | ONE.ie | 75 | Technical deep-dive on 'anthropic skills jar' with Anthropic Skills examples, ON |
| ai agents course free with certificate | D | onlineoptimisers.ai | 75 | Tutorial series entry for 'ai agents course free with certificate' with 5-step b |
| marketing plan for cryptocurrency | E | onlineoptimisers.ai | 55 | Agency playbook on 'marketing plan for cryptocurrency' with before/after case da |
| on-chain seo | E | onlineoptimisers.ai | 55 | Agency playbook on 'on-chain SEO' with before/after case data, AEO answer capsul |
| 7 streams of income to become a millionaire | F | ONE.ie | 80 | Escrow / payments primer on '7 streams of income to become a millionaire' with s |
| acp escrow | F | ONE.ie | 80 | Escrow / payments primer on 'ACP escrow' with smart contract code, rev-share mat |
| agent invoice smart contract | F | ONE.ie | 80 | Escrow / payments primer on 'agent invoice smart contract' with smart contract c |
| x402 escrow | F | ONE.ie | 80 | Escrow / payments primer on 'x402 escrow' with smart contract code, rev-share ma |
| agent identity blueprint | G | ONE.ie | 55 | Trust primitives explainer for 'agent identity blueprint' with DID/reputation ar |
| verifiable agent outputs | G | ONE.ie | 40 | Trust primitives explainer for 'verifiable agent outputs' with DID/reputation ar |
| claude code plan mode save plan | I | onlineoptimisers.ai | 55 | Claude Code tutorial on 'claude code plan mode save plan' with code examples, 3 |
| claude code plan mode system prompt | I | onlineoptimisers.ai | 55 | Claude Code tutorial on 'claude code plan mode system prompt' with code examples |
| claude code plan mode tutorial | I | onlineoptimisers.ai | 55 | Claude Code tutorial on 'claude code plan mode tutorial' with code examples, 3 u |
| claude code plan mode vs edit mode | I | onlineoptimisers.ai | 55 | Claude Code tutorial on 'claude code plan mode vs edit mode' with code examples, |
| claude code worktree command | I | onlineoptimisers.ai | 55 | Claude Code tutorial on 'claude code worktree command' with code examples, 3 use |
| claude code worktree support | I | onlineoptimisers.ai | 55 | Claude Code tutorial on 'claude code worktree support' with code examples, 3 use |
| why is it called vibe coding | I | onlineoptimisers.ai | 55 | Claude Code tutorial on 'why is it called vibe coding' with code examples, 3 use |
| train llm on content | J | onlineoptimisers.ai | 55 | LLM-training-data SEO playbook for 'train LLM on content', covering arXiv + Wiki |
| agent algo coin | K | agent-launch.ai | 55 | Launchpad guide on 'agent algo coin' with bonding curve math, legal language pla |
| llm router langchain | L | ONE.ie | 55 | Inference routing benchmark for 'llm router langchain' with TypeDB router story, |
| llm router nvidia | L | ONE.ie | 60 | Inference routing benchmark for 'llm router nvidia' with TypeDB router story, co |
| routellm ai | L | ONE.ie | 60 | Inference routing benchmark for 'routellm ai' with TypeDB router story, cost/lat |
| routellm ai free | L | ONE.ie | 55 | Inference routing benchmark for 'routellm ai free' with TypeDB router story, cos |
| semantic router github | L | ONE.ie | 60 | Inference routing benchmark for 'semantic router github' with TypeDB router stor |
| semantic router langchain | L | ONE.ie | 55 | Inference routing benchmark for 'semantic router langchain' with TypeDB router s |
| nano claw with ollama | ONE | ONE.ie | 80 | ONE.ie substrate explainer for 'nano claw with ollama', benchmark dashboard link |
| typedb ai | ONE | ONE.ie | 55 | ONE.ie substrate explainer for 'TypeDB AI', benchmark dashboard link, architectu |
| ai seo company coalition | OO | onlineoptimisers.ai | 60 | Local / agency landing page for 'ai seo company coalition' with case studies, Lo |
| ai visibility audit agency | OO | onlineoptimisers.ai | 55 | Local / agency landing page for 'AI visibility audit agency' with case studies, |
| rankar ai | OO | onlineoptimisers.ai | 55 | Local / agency landing page for 'rankar ai' with case studies, LocalBusiness + S |
| custom ai chatbot agency | OOI | oo-integrations.ai | 55 | Enterprise AI consulting page for 'custom AI chatbot agency' with whitepaper CTA |
No invented data, no optimistic estimates. Real gaps that could affect strategy decisions.
| Blind Spot | What's Missing | Impact |
|---|---|---|
| Gemini-2.5-flash overloaded (503 on launch day) | Model at capacity on 2026-04-30. Google AI Studio API key is in .env and working. Gemma-3-27b (same API, open-weight Google model) used as interim. Will re-run full 50 with Gemini-proper when overload clears. | PARTIAL |
| Full page word counts not scraped | We used meta description word count as proxy (avg 22-27 words). Full page scraping via Firecrawl is P1 research task. Gap analysis is directional, not precise. | MEDIUM |
| No real Ahrefs DR for incumbents | Domain authority estimated by tier (tier1=92, AI-infra=82, vendor=55). Real DR from DataForSEO Backlinks Bulk would sharpen "rankable" calls. | MEDIUM |
| No rank tracking baseline | We know what we SHOULD rank for. We don't know where we currently rank. No progress measurement without a baseline position per P1 keyword. | HIGH |
| No brand mention monitoring | pheromone routing, ONE.ie, IoA, AGNTCY on Reddit / HN / Discord / X not monitored. Missing link opportunities and conversations. | MEDIUM |
| Brand collision - onlineoptimisers.com | Q43 "Online Optimisers review" - Google's Gemma cited onlineoptimisers.com which forwards to onlineoptimisers.eu - a completely separate EU analytics company "Online Optimizers" (with a 'z'). British vs American spelling distinction is subtle. Buy .com or proceed with .ai as the primary brand to sidestep the collision. | LOW-MED |
| Backlink quality not per-URL scored | 2,481 URLs identified, no Ahrefs UR per URL. Volume large enough that 20% hit rate is excellent. | LOW |
| Perplexity Q46-50 RESOLVED | Re-run 2026-04-30. Q48 (GEO optimization) = 0 citations - VIRGIN. Q49 (Cloudflare Workers) = 0 - VIRGIN. Q50 = fastbots.ai cited (ONE.ie gap). | RESOLVED |
| Priority | Research task | Why | Cost |
|---|---|---|---|
| IN PROGRESS | Gemini-2.5-flash full baseline (50 queries) | Gemma-3-27b interim data collected. Gemini-proper baseline needed for Google AI Overviews data. Re-run when gemini-2.5-flash overload clears (typically 24-48h after peak load). | $0 (free tier) |
| DONE | Perplexity top-up + re-run Q46-50 | Completed 2026-04-30. Q48+Q49 confirmed VIRGIN. Q50 = fastbots.ai. | ~$5 |
| P1 | Full-page word count scrape (Firecrawl on top-10 SERP per P1 kw) | Real content gap numbers vs the desc-word-count proxy. Sharpens P1 prioritisation. | ~$15 |
| DONE | Rank tracking baseline (25 P1 keywords, live SERP) | Completed 2026-04-30. 0/25 ranked (brand new domains - expected baseline). Incumbents mapped. 2 keywords removed (tool router + agent treasury = wrong SERP). See SERP Intelligence section above. | ~$0.05 done |
| DONE | Real DR for SERP incumbents (DataForSEO Backlinks Bulk) | Completed 2026-04-30. tidio.com 19,464 | langchain.com 16,407 | fetch.ai 5,069 | one.ie 42 | agent-launch.ai 0. 120x gap confirmed. See domain authority chart above. | ~$0.02 done |
| P2 | Expand LLM baseline to 150 queries | 50 queries is directional. 150 gives statistical confidence on citation rates. | ~$15 |
| P2 | GitHub star velocity tracking (ONE SDK repos) | Stars = strong LLM training signal. Cited repos compound in AI answers. Free via GitHub API. | $0 |
| ACTIVE | DataForSEO LLM Mentions API - already live | Confirmed active in DataForSEO dashboard (renews 23.05.2026). No action needed. Full production-grade citation tracking available. LLM Mentions endpoint to be wired into monthly cron. | Paid, active |