The numbers that matter
State of AI search — Q1 2026
60%
of searches are zero-click. The answer appears before a link is ever visited.
AEO.dev / Semrush 2025
4B
monthly ChatGPT visits. Now among top 10 websites globally.
AEO.dev market data
50%
Google traffic decline projected by 2028 if AI search trajectory continues.
Semrush / AEO.dev
32%
of AI search citations come from Reddit — the single most-cited domain.
AEO.dev research
0.65
correlation between Google page-1 rankings and LLM brand mentions.
AEO.dev / 10K+ prompts
40%
enterprise apps will embed AI agents by EOY 2026 (up from 5% in 2025).
Gartner 2026
Essential reading
Best sources, summarized
01
Essential
The most organized public resource on AEO. Community-driven, GitHub-backed, and structured around the full implementation stack: what it is, how it differs from SEO, optimization strategies, technical implementation, authority building, and measurement. The research section is particularly strong — based on testing 10,000+ AI search prompts to rank which optimization methods actually improve citation visibility. The checklist alone is worth the visit.
- Quotes and statistics are the highest-impact content moves — 30–40% visibility improvement measured
- Reddit is the #1 cited domain in AI search (~32%). Wikipedia has 5× LLM training data weight
- The three-pillar framework (on-page, earned authority, community) is the right mental model
- Traditional SEO still matters — 0.65 correlation with LLM mentions means SEO work supports AEO
- Best starting point: aeo.dev/checklist — prioritized implementation in four weekly phases
02
Essential
One of the most practically useful AEO guides available. Strong on the mechanical difference between how AI systems process content vs. traditional search — the three-step flow: intent parsing, candidate retrieval, selection. The key insight on query length is actionable: LLM prompts average 13 words vs. 3 for traditional search, which means content structured for 3-word keywords misses the actual query patterns AI systems handle.
- Average LLM prompt is 13 words. Write headings and content for natural language questions, not keywords
- Pages with exact or close language matches to query phrasing are significantly more likely to be cited
- AEO measurement requires tracking citation share and brand mentions — not rankings or CTR
- Teams get the most lift when treating AEO as a system targeting middle-funnel decision queries
- Answer engines evaluate for clarity, directness, and trust signals — not just ranking position
03
Deep read
The most thorough treatment of AEO as a business discipline. The NerdWallet case study is the most cited evidence that AEO works even as traditional traffic declines — 35% revenue growth despite a 20% decrease in site traffic, by ensuring content and brand expertise still reached consumers through AI-generated answers. Good on the architectural distinction between answer engines (NLP-driven intent parsing) and search engines (keyword matching).
- NerdWallet case: 35% revenue growth with 20% traffic drop by prioritizing AI-generated answer presence
- Voice commerce projected at $80B annually — agents integrating with business systems for transactions
- AEO is not a replacement for SEO; it's a necessary extension ensuring brand visibility as search modes shift
- 99% of URLs shown in Google AI Mode also appear in top 20 organic results — SEO correlation confirmed
- Best for: building the executive case for AEO investment alongside traditional SEO
04
Practical
Profound's playbook is the most execution-focused AEO resource available — step-by-step process for building AEO around a specific brand. It's notably candid about AI visibility tools (they sell one), but the methodology is sound: audit what AI currently says about your category, identify which sources dominate citations, reverse-engineer their content patterns, then produce content structured to replace them. The competitive intelligence step is underrated.
- Start by auditing AI citations in your category — see who is currently being cited and why
- A CPG brand used competitive intelligence to find a rival was using custom press pages to dominate AI responses
- Content for AEO: clear headings, bullet points, comparison tables, FAQ sections — not prose essays
- Distribution matters as much as creation: get content onto platforms AI models train from
- Measurement: track brand mentions in AI responses across platforms, not just traffic referral
05
Technical
The most concise developer-focused AEO primer. Strong practical framing: "The easiest time to optimize for AEO was 2025. The second easiest time is today." The core insight is architectural — SEO rewards findability, AEO rewards extractability. A page that ranks #3 on Google but buries its answer in paragraph five won't get cited by AI even if Google loves it. Includes a useful 30-minute quick-start that delivers genuine impact before any strategy work.
- AEO is simpler than SEO — fewer competitors, clearer rules, higher ROI right now
- 60% of indie creator sites are invisible to AI crawlers — most optimization hasn't happened yet
- 30-minute quickstart: robots.txt + llms.txt + FAQPage schema = highest-impact actions available
- Content ranking on Google does not automatically appear in AI search results — separate optimization needed
- Each H2 should be a complete, self-contained answer — extractable verbatim without context
06
Workflow
The most operations-focused AEO resource — written for marketing teams that need workflows, ownership assignments, and checklists rather than strategy theory. HubSpot's credibility comes from a year of internal testing and experimentation. The key framing: AEO complements SEO and inbound marketing, doesn't replace them. Good at translating AEO principles into repeatable content production processes with clear roles and handoffs.
- AEO is an ongoing practice, not a launch project — unlike SEO which evolved over decades, AEO is moving in months
- Map user questions → structure content for quick answers → add schema → track visibility
- 79% of AI search users believe AI provides a better experience than traditional search
- HubSpot's AEO Grader is a free tool for auditing current AI visibility — useful baseline
- Best for: teams building repeatable content workflows, not one-off optimization projects
07
Action plan
The most structured 90-day AEO implementation roadmap available. Distinct from their general playbook — this one is phase-gated with specific actions per 30-day window. The key insight is sequencing: start with existing pages that already rank (quick wins with established authority), not from scratch. The cross-functional role breakdown is the most complete version of "who does what" in an AEO program.
- Days 1–30: prioritize pages already ranking but lacking AEO structure — fastest ROI
- Days 31–60: implement Q&A headings, schema, FAQ blocks; add answer summaries at article tops
- Days 61–90: distribute optimized content to authoritative platforms; launch citation monitoring
- Roles needed: content strategist (intent mapping), editor (atomic formatting), SEO/dev (schema implementation)
- Research shows 99% of AI Mode citations come from pages already in top 20 organic results
AEO for autonomous agents specifically
The AEO layer above GEO
Most AEO resources above are really GEO with an updated name — optimizing for AI-generated answers in chat interfaces. The AEO Observatory at wwjd.dev/auto studies what happens when the AI is an agent — not a search interface, but a system that browses, acts, and propagates decisions without a human in the loop.
Agent-specific AEO extends the GEO stack with surfaces that only matter when the reader is an autonomous system: instruction-layer HTML attributes, structured notification payloads, memory instructions, and behavioral observation. This is the research frontier. The resources above don't cover it — it's being mapped here.
→ wwjd.dev/auto — the live observatory
→ aeo_setup — implementation including the AEO-specific layer
→ @fingogh — ongoing research and field notes