Chapter 1 Audience Strategy (Who to Target): Five Marketing Audiences and Six Optimization Domains · IDAEO: Marketing in the Age of AI
Audience Strategy (Who to Target): Five Marketing Audiences and Six Optimization Domains
This chapter answersWhat do the different Marketing Audiences come to do, and which Optimization Domains do they map to?
Why it comes hereIdentify audiences and functions first; only then can you read the journey numbers.
Figure 1-1 Panorama of the Five Marketing Audiences: Functions, Optimization Domains, and Measured Results
Read from top to bottom: Audience Strategy, Marketing Audiences, functions, main related domains, number of visits, visitor types, and the top four; each name has its English below it. The main related domains list abbreviations and full names, with related terms for the same domain below. Counts are by function. GEO appears in the domain legend, with no count attached. The “makes AI Overviews” cell is led by AIO and linked to AEO: AIO focuses on search engine AI Overviews, and AEO gets AI answers to use you. For Humans, the functions list “Organic Traffic” and “Local Search” side by side; Automated Bots provide “Automated Information Services.” AEO also supports AI real-time reads, AI Agents, and Automated Information Services. AEO is an Optimization Domain focused on getting AI answers to use content. GEO is an Optimization Domain focused on making content more visible in generative answers. Both involve the answer stage, and their scopes overlap.
View as text
Read top to bottom: Audience Strategy, Marketing Audiences, function, Optimization Domains, visit counts, visitor types, and the top four.
① Audience For whom Five Marketing Audiences: Functions, Optimization Domains, Measured Results It Is Not Only People Who Read You: Five Marketing Audiences, Each with Its Own Way of Reading Audiences Five Audiences Humans Humans Search Engine Crawlers Search Engine Crawlers AI Crawlers AI Crawlers AI Agent AI Agents Automated Bots Automated Bots Functions What It Comes to Do OrganicTraffic Organic Traffic Local Search Local Search Build Search Index Build Search Index Generate AI Overviews Generate AI Overviews Copy Back for Training Model Training Build AI Index Build AI Index Real-TimeRetrieval Real-Time Retrieval Act for Users Address Lookups Act on Behalf of Users Automated Information Services Automated Information Services Domains Six Domains Visits 3/3–10/8 Visitor Types 3 types Search, AI, Maps 15 types 8 types 7 types 13 types 13+ types 20+ types Most Frequent Top four by count 268,000 20,000 10,000 9,388 24.62M 12.05M 3.41M 510,000 24.48M 11.58M 10.6M 5.45M P 2.82M 760,000 220,000 130,000 640,000 11,000 2,494 1,624 Browsing 17.95M TollBit 130,000 mcpbeat 7,925 5,683 Golemreach AI Tools 7.08M Tracking 150,000 SEO crawlers 3.51M Previews 170,000 ID EO Original framework © 2026 Audience Strategy Marketing Audiences Functions Optimization Domains Visits Visitor Types Top Platforms Organic Traffic (from 8/26) Local Search Same Search Engine Crawlers, shared by two functions OrganicTraffic Result of the Six Optimizations MEO Map Engine Optimization Local SEO GBP Optimization SEO SearchEngine Optimization SXO Technical SEO AIO AI Overview Optimization AEO Answer Engine Optimization SGE Optimization LLMO LLM Optimization LLM SEO S4 function Builds AI index (not GEO) GEO: generative answer visibility See glossary AEO Answer Engine Optimization Answer Engine Optimization VSO (Voice Search), AAO, AXO, AAIO (Optimization for AI Agents) 309,826 41.26M 52.2M 3.99M 0.65M 18.09M 7.23M Map Engine Optimization Search Engine Optimization AI Overview Optimization Large Language Model Optimization Indexing function, not GEO Answer Engine Optimization Tracking
Terms from the same field are listed directly under the main name according to how they are used. AIO is linked to AEO; the industry uses AIO to mean “AI Overview Optimization” or as a general term for “AI optimization,” and both correspond to getting AI answers to use your content. In IDAEO, AIO refers to optimizing search engine AI Overviews. ASO, CRO, and SMO belong to traditional digital marketing and work alongside the Six Optimization Domains.
How to readRead top to bottom: Audience Strategy, Marketing Audiences, function, Optimization Domains, visit counts, visitor types, and the top four.
Figure 1-2 AI Traffic Classification Tree: How Do You Tell Which Type of AI Traffic an AI Request Belongs To?
View as text
Start from the request at the top and read down through the classification basis and function; the end point is which of the Marketing Audiences it belongs to.
How do you decide which category of AI traffic an AI request belongs to? (2026/3/3–10/8) Rounded box = start/end; diamond = decision; rectangle = result; dashed box = Black Box (inside AI Platforms); arrow thickness roughly follows volume. A request comes in 197 million requests Is it an AI/search request? (AI use + browsing) No: 80.5M Type: AI tools 7.08 million Type: Tracking in Progress 153,000 Type: all other traffic 73.27 million Yes, 116 million Why is it here? (AI traffic type) Type: AI Crawlers·training S3 Train 52.2 million requests Anthropic, Meta, etc. Type: search crawlers S4 Index 41.26 million requests Search index + AI Overviews Type: AI index crawlers S4 Index 3.99 million requests Builds AI index (our AI Search, etc.) Type: real-time crawlers S5 Ask 651,000 requests Real-time (OpenAI 635,000) Type: task-tool crawlers S8 Act 144,000 requests Acts for people (logged) Type: AI Agents browsing S8 Act 17.95 million requests AI Agents (automated browsing) In-platform: Tracking Black Box: inside AI Platforms After training and indexing, use in AI answers: Tracking in Progress Black Box: inside AI Platforms Task process: Tracking Was it used by search or AI? Google/Bing provide search & AI data Others Black Box: inside AI Platforms Search/AI use on other platforms: Tracking in Progress Yes (Google, Bing) Type: S4 Index 657,000 pages Bing + Google (September) Type: S6 Mention about 8.22 million times Search impressions (approx.) Type: S7 Answer 31,711 times Copilot (7/22–10/8) Did customers visit? No Type: Tracking in Progress Referral sources still being tracked Yes, 309,826 Type: S9 Visit (Organic Traffic) 309,826 (referral source records) Google 267,734・Microsoft 20,053・Yahoo Japan 10,278 1. Period: total site traffic 2026/2/12–10/8; logged AI categories 2026/3/3–10/8; AI Agents automated browsing 2026/8/26–10/8. 2. “Yes” = identifiable AI/search use, 116,203,857 requests; “No” = Automated Information Services 7,231,923 requests + all other traffic 73,266,746 requests. 3. Indexed pages are approximate figures for September 2026 (Bing 375,000 + Google 282,000); search result impressions about 8.22 million (Google + Bing); AI citations are Copilot, 2026/7/22–10/8. 4. Organic Traffic of 309,826 is from referral source records; its unit of measurement differs from indexed pages, search result impressions, and AI citations. Meta 9,388·DuckDuckGo 1,049·Baidu 713·other sources 611 ID EO
| Item | Value |
|---|---|
| Type: AI tools | 7.08 million |
| Type: Tracking in Progress | 153,000 |
| Type: all other traffic | 73.27 million |
| Type: AI Crawlers·training | 52.2 million |
| Type: search crawlers | 41.26 million |
| Type: AI index crawlers | 3.99 million |
| Type: real-time crawlers | 651,000 |
| Type: task-tool crawlers | 144,000 |
| Type: AI Agents browsing | 17.95 million |
How to readStart from the request at the top and read down through the classification basis and function; the end point is which of the Marketing Audiences it belongs to.
Figure 1-3 AI Platforms and Crawlers Family Tree: Which Companies and Which Crawlers Sit Under Each Function
Extended figures: Figure 1-2, “AI Traffic Classification Tree,” determines which type of AI traffic an AI request belongs to; Figure 1-3, “AI Platforms and Crawlers Family Tree,” lists the AI Platforms and crawlers under each function.
View as text
Read from root to branch: function, AI Platforms, and crawlers; the same company on different branches means it takes part in different functions.
Crawler Family Tree 2026/3/3–10/8 · 123 million AI requests → purpose → AI Platforms → crawlers/automated browsing (60) · branch width = √requests Root Logged + separate browsing Layer 1 · Purpose 7 categories, by function & audience Layer 2 · AI Platforms Dot = brand color/gray Layer 3 · Crawlers/automated browsing (60 entries) Capsule = name + requests; gray = function; sorted by requests AI requests (all) 123 million Unit: number of requests Copies to train S3 52.2M 42.3% of total AI Crawlers Builds search index S4 41.26M 33.4% of total Search crawlers + AI Overviews Builds AI index S4 3.99M 3.2% of total AI Crawlers Someone asks AI S5 651,000 0.5% of total Real-time retrieval (AI Crawlers) Acts for customers S8 18.09M 14.7% of total AI Agents (incl. automated browsing) AI Tools 7.08M 5.7% of total Automated Bots · Info Services Tracking 153,000 0.1% of total Automated Bots · Info Services Anthropic Meta OpenAI Amazon Tracking Common Crawl Microsoft Google Yandex ByteDance Sogou Tracking DuckDuckGo Baidu Brave Yahoo Japan Apple Our AI Search OpenAI Perplexity Parallel Huawei Linkup Anthropic You.com Tracking Kagi OpenAI Perplexity Anthropic DuckDuckGo Google Tracking Mistral Separately TollBit mcpbeat Golemreach Tracking Andi AgentTools-Cloud Tracking Google DeepSeek Tracking Anthropic Cohere NICT ROIS ClaudeBot 24.48M Copies back for training meta-externalagent 11.58M Copies back for training GPTBot 10.6M Copies back for training Amazonbot 5.45M Copies back for training Reflectionbot 64,000 Copies back for training Info service crawlers 307 Copies back for training Other browsers 39 Copies back for training CCBot 19,000 Copies back for training bingbot 12.05M Search index + AI Overviews Googlebot 3.41M Search index + AI Overviews YandexBot 511,000 Search index + AI Overviews Bytespider 256,000 Search index + AI Overviews Sogou 253,000 Search index + AI Overviews Other crawlers 31,000 Search index + AI Overviews DuckDuckBot 12,000 Search index + AI Overviews Baiduspider 4,630 Search index + AI Overviews Bravebot 39 Search index + AI Overviews Y!J-DLC 30 Search index + AI Overviews Applebot 24.62M Search index + AI Overviews AI Search 2.81M Builds AI index AI Search External 5,918 Builds AI index OAI-SearchBot 761,000 Builds AI index PerplexityBot 218,000 Builds AI index ShapBot 126,000 Builds AI index PetalBot 114,000 Search index + AI Overviews LinkupBot 60,000 Builds AI index Claude-SearchBot 11,000 Builds AI index YouBot 596 Search index + AI Overviews Googlebot-Video 5 Search index + AI Overviews Kagibot 4 Search index + AI Overviews ChatGPT-User 635,000 Real-Time Retrieval Perplexity-User 11,000 Real-Time Retrieval Claude-User 2,494 Real-Time Retrieval DuckAssistBot 1,624 Real-Time Retrieval Google NotebookLM 464 Real-Time Retrieval Gemini DeepResearch 10 Real-Time Retrieval Python-urllib 19 Real-Time Retrieval TABINOTE-ArrivalAudit 2 Real-Time Retrieval curl 2 Real-Time Retrieval HatchFeed 1 Real-Time Retrieval MIN-SourceReview 1 Real-Time Retrieval Meta-WebIndexer 1 Real-Time Retrieval MistralAI-User 6 Real-Time Retrieval AI Agents (browsing) 17.95M Browsing (separate) tollbot 126,000 Errands, address lookups mcpbeat 7,925 Errands, address lookups GolemreachTrustBot 5,683 Errands, address lookups Real-browser AI Agents 5,102 Errands, address lookups (separate) Andi-AgentSurfaceProbe 1,956 Errands, address lookups agent-tools.cloud-crawler 1,866 Errands, address lookups AI tools 7.08 million AI Tools GoogleOther 142,000 Tracking in Progress GoogleOther-Image 34 Tracking in Progress DeepSeek 9,551 Tracking in Progress HanaleiBot 574 Tracking in Progress anthropic-ai 27 Tracking in Progress Claude-Web 21 Tracking in Progress CohereBot 30 Tracking in Progress ICC Crawler 30 Tracking in Progress Cotoyogi 2 Tracking in Progress Logged 105,487,618 requests (2026/3/3–10/8) + AI Agents automated browsing 17,948,162 requests (2026/8/26–10/8) = tree root 123,435,780 requests. Automated Bots = AI tools 7,079,310 requests + Tracking in Progress 152,613 requests; the Search Engine Crawlers' search indexing and AI Overview generation come from the same batch of requests. Branch thickness follows the original figure, scaled to the square root of request counts (at least 1.5px); each row is attached only once. Capsule values are rounded in units of ten thousand / one hundred million. ID EO
| Item | Value |
|---|---|
| ClaudeBot | 24.48M |
| meta-externalagent | 11.58M |
| GPTBot | 10.6M |
| Amazonbot | 5.45M |
| Reflectionbot | 64,000 |
| Info service crawlers | 307 |
| Other browsers | 39 |
| CCBot | 19,000 |
| bingbot | 12.05M |
| Googlebot | 3.41M |
| YandexBot | 511,000 |
| Bytespider | 256,000 |
| Sogou | 253,000 |
| Other crawlers | 31,000 |
| DuckDuckBot | 12,000 |
| Baiduspider | 4,630 |
| Bravebot | 39 |
| Y!J-DLC | 30 |
| Applebot | 24.62M |
| AI Search | 2.81M |
| AI Search External | 5,918 |
| OAI-SearchBot | 761,000 |
| PerplexityBot | 218,000 |
| ShapBot | 126,000 |
| PetalBot | 114,000 |
| LinkupBot | 60,000 |
| Claude-SearchBot | 11,000 |
| YouBot | 596 |
| Googlebot-Video | 5 |
| Kagibot | 4 |
| ChatGPT-User | 635,000 |
| Perplexity-User | 11,000 |
| Claude-User | 2,494 |
| DuckAssistBot | 1,624 |
| Google NotebookLM | 464 |
| Gemini DeepResearch | 10 |
| Python-urllib | 19 |
| TABINOTE-ArrivalAudit | 2 |
| curl | 2 |
| HatchFeed | 1 |
| MIN-SourceReview | 1 |
| Meta-WebIndexer | 1 |
| MistralAI-User | 6 |
| AI Agents (browsing) | 17.95M |
| tollbot | 126,000 |
| mcpbeat | 7,925 |
| GolemreachTrustBot | 5,683 |
| Real-browser AI Agents | 5,102 |
| Andi-AgentSurfaceProbe | 1,956 |
| agent-tools.cloud-crawler | 1,866 |
| AI tools | 7.08M |
| GoogleOther | 142,000 |
| GoogleOther-Image | 34 |
| DeepSeek | 9,551 |
| HanaleiBot | 574 |
| anthropic-ai | 27 |
| Claude-Web | 21 |
| CohereBot | 30 |
| ICC Crawler | 30 |
| Cotoyogi | 2 |
How to readRead from root to branch: function, AI Platforms, and crawlers; the same company on different branches means it takes part in different functions.
Figure 1-4 IDAEO Traffic Atlas: Five Traffic Categories, Each Family, and Every Crawler
View as text
From left to right are the five traffic categories, with each category's total at the top of its column; within a column, rows are grouped by family, one crawler per row with its visit count, and a blank means no traffic. In the AI Crawlers column, the dot on the three-position track shows which journey step that crawler sits at. This figure follows each company's official documentation (checked Oct 14, 2026), which places Applebot, PetalBot, YouBot, and Kagibot under Search Engine Crawlers, the same classification used in the main text. Data period Mar 3, 2026–Oct 8; ＊ marks records counted separately.
How to readFrom left to right are the five traffic categories, with each category's total at the top of its column; within a column, rows are grouped by family, one crawler per row with its visit count, and a blank means no traffic. In the AI Crawlers column, the dot on the three-position track shows which journey step that crawler sits at. This figure follows each company's official documentation (checked Oct 14, 2026), which places Applebot, PetalBot, YouBot, and Kagibot under Search Engine Crawlers, the same classification used in the main text. Data period Mar 3, 2026–Oct 8; ＊ marks records counted separately.
Key Points of This Chapter
| Claim | Why |
|---|---|
| The Five Marketing Audiences are Humans, Search Engine Crawlers, AI Crawlers, AI Agents, and Automated Bots. | Each type visits for a different purpose; only by measuring them separately can you see the marketing value clearly. |
| Classify AI traffic by function first, then look at the AI Platforms. | The same AI platform may run training, indexing, and real-time retrieval at the same time; function says more about what the content is read for. |
| Visits by AI Agents have been observed, and their scale is still being tracked. | AI is moving from “answering” to “acting for customers,” and site visitors are extending from crawlers that read content to AI Agents that carry out tasks. Measured AI Agent Visits have already reached 18.09 million, of which automated browsing accounts for 17.95 million (counted from Aug 26), classified by behavior, and named AI Agents account for 144,000, showing the scale of automated visits. |
| The Six Optimization Domains are SEO, AIO, AEO, GEO, LLMO, and MEO. SEO builds the search index, AIO focuses on search engine AI Overviews, and AEO gets AI answers to use your content. GEO improves visibility in generative answers, LLMO corresponds to copying back for training, and MEO corresponds to Local Search. AEO and GEO both operate at the answer stage, and AIO is linked to AEO. Building the AI index is the name of an S4 function, not GEO. | AI Overviews are also one form of AI answers using content, so AIO is subordinate to AEO. AEO and GEO both operate at the answer stage, so the domains overlap. |
| AI, search engines, and Automated Bots combined: 105.5 million (since Mar 3; these are platform-scale figures and cannot be compared directly with a single merchant's website). Organic Traffic from Humans: 310,000 (since Aug 26). Requests before Aug 24 may include duplicates. | The two figures cover different periods and each shows scale. Content must be acquired before it can support later use, so AI reading is the foundation of marketing. |
| AI Crawlers read content 52.2 million times for training and 3.99 million times for building the AI index (including 2.82 million by the site's own Cloudflare AI Search), and made 650,000 real-time reads. | Bulk reads and question-triggered reads serve different purposes, so the counts must be read together with function. |
| The Optimization Domains in Figure 1-1, “Panorama of the Five Marketing Audiences,” list the abbreviation and its full name, with terms for the same domain listed below. | Different common names are gathered in the same cell, so IDAEO Learners can map industry terms back to the same function. |
Terms
- Five Marketing Audiences
- Humans, Search Engine Crawlers, AI Crawlers, AI Agents, and Automated Bots.
- Functions
- The work each of the Marketing Audiences performs, such as copying back for training, indexing, and real-time retrieval.
- AI Agent
- AI that acts for customers.
- Automated Information Services
- Services such as information organization, analysis, and web page previews provided by Automated Bots.
- Organic Traffic
- The number of times customers click through to the website from search or AI.
- Local Search
- Customers looking for merchants or services by location and nearby needs.
- AIO
- In this book, AIO refers only to AI Overview Optimization for search engines. When AIO is used in the market as a general term for AI Optimization, it spans multiple domains. That broader usage cannot be placed entirely under this book's AEO.
- Tracking in Progress
- IDAEO keeps accumulating measurement results and will publish them later.
FAQ
Q: What kinds of AI Crawlers are there?
A: There are three types by function: copying back for training, building the AI index, and real-time retrieval. Crawling for training is intended to let models accumulate information. Crawling for indexing is intended to bring content into sources for later use. Real-time retrieval responds to a customer's question in the moment. A crawl does not mean the content has been used for training or indexed. Look at the function first, then at which company, because the same AI Platform may do several jobs. For how to tell them apart, see Figure 1-2, “AI Traffic Classification Tree”; for every company and crawler, see Figure 1-3, “AI Platforms and Crawlers Family Tree”; for which companies appear in each of the five traffic categories and how many visits every crawler made, see Figure 1-4, “IDAEO Traffic Atlas.” AI Agents act for customers and are counted separately from AI Crawlers.
Q: What is the difference between AI Crawlers and AI Agents?
A: AI Crawlers mainly acquire information; AI Agents mainly act for customers. Performance is measured separately: the number of times content was acquired, and the results of acting on customers' behalf.
Q: Visits by AI Agents have been observed. Why keep tracking their scale?
A: AI is moving from providing answers to acting on customers' behalf. Customers hand searches, bookings, and other tasks to AI, so the AI that arrives at the site has to understand the information and complete the job; merchant marketing therefore has to support content use and acting on customers' behalf.
Q: Is the industry's AIO the same as IDAEO's AIO?
A: It depends on the usage. In this book, AIO refers only to AI Overview Optimization for search engines. When AIO is used in the market as a general term for AI Optimization, it spans multiple domains. That broader usage cannot be placed entirely under this book's AEO.
Q: How do you decide which category of AI traffic an AI request belongs to?
A: Following Figure 1-2, “AI Traffic Classification Tree,” look at what this visit does: copying back for training, indexing, real-time retrieval, or acting for customers. Function determines the category, because the same AI Platforms may perform several kinds of work.
Q: Why do the 17.95 million automated browsing visits count as AI Agents?
A: This book counts the 17.95 million automated browsing visits as AI Agents and labels them separately as automated browsing. Programs operating a browser show only that execution is automated; they do not prove that a customer requested the task. Classification follows the measured function. Verifiable visits from named AI Agents total 144,000.
Q: Where are the Local Search numbers?
A: Local Search is still being tracked and will be published later; Figure 1-1, “Panorama of the Five Marketing Audiences,” currently lists Organic Traffic. Local Search and Organic Traffic use different statistical definitions, so the results are presented separately.
Q: How does AEO optimization differ from SEO?
A: SEO focuses on search engine indexing and search presentation; AEO gets AI answers to use your content, covers AIO's overview optimization, and supports AI real-time retrieval, AI Agents, and Automated Information Services. Search presentation and AI answers reach people differently, so results must be read together with function.
Q: What do GEO and LLMO each cover?
A: LLMO focuses on model training. A crawler reading content for training does not mean the model has completed training. GEO is the optimization domain that improves content visibility in generative answers, consistent with Aggarwal et al. (2024), cited at the end of the book. Building the AI index is an S4 function. The 3.99 million reads in Figure 1-1 are reported under that function name.
Q: Where do conversion rate optimization and social media optimization go?
A: ASO, CRO, and SMO belong to traditional digital marketing and serve people. This work improves how people reach and choose, and can work alongside the Six Optimization Domains.
Q: Which kind of crawler reads each of the four ways AI mentions you?
A: ① is read by AI Crawlers for training, at S3 Train. ②③④ are read by Search Engine Crawlers and AI Crawlers, at S4 Index and S5 Ask. See Figure 2-5.
This Chapter in Three Sentences
- A website has five types of readers, each visiting for a different reason.
- When examining AI traffic, first look at what it came to do, then at which company it comes from.
- Retrieval only means the content was obtained. It does not mean it was cited or that a customer visited.
Three Things You Can Do This Week
- Clearly list your business name, address, opening hours, and services on your website.
- Open your website's visitor logs and look separately at visits from Humans, search engines, and AI.
- Record AI crawls separately by purpose: training, indexing, and real-time retrieval.
Next ChapterIn the next chapter, you will learn the diagnostic checkpoints for AI crawls, AI mentions, and customer visits.
Self-CheckBased on one record of “someone asked AI, and AI immediately came to read the merchant page,” identify which of the Marketing Audiences it belongs to, its function, and which of the Optimization Domains it maps to, then explain what the merchant must support when AI acts for customers.
Free to download, updated monthly
Cite this book
IDAEO: Marketing in the Age of AI · 林毅祥（TK Lin） · 1st edition, 9th revision · · https://km.idaeo.ai/idaeo/textbook?lang=en · CC BY 4.0