How to Use This Encyclopedia

Reading paths

Five paths through the document. Pick the one that matches your operating context.

Linear path. Start at Chapter 1 and read through Chapter 47 in order. Build the operating system from organic foundations to the paid stack. Best if you are new to GEO or want the full mental model. Time investment: a focused weekend, then the work itself.

Modular path (organic). Use the table of contents and the 14-point readiness checklist in What Agent-Ready Means to identify which pillar is broken, then jump to the relevant chapter. Best if you already have an SEO foundation and want to fix specific gaps. Time investment: 30-60 minutes per gap.

Zero-to-one paid. You have never run ChatGPT Ads. Read chapters 29-34 first to understand the product itself (where ads appear, who sees them, how OpenAI matches an ad to a conversation, the privacy model). Then chapters 35-40 for the measurement stack (Ads Manager, JavaScript Pixel, Conversions API, event taxonomy, dedup). Finally chapters 41-47 for the Advertiser API.

Integration path. You are wiring up the pixel and Conversions API on a Shopify store right now.

Chapter 36. Pixel install snippet, exact <head> placement, automatic oppref capture, and the CSP allowlist for security-conscious sites.
Chapter 38. Pick the events you need. For Shopify the minimum stack is page_viewed, contents_viewed, items_added, checkout_started, order_created.
Chapter 39. The data shapes. Get amount denomination right (integers in lowest currency unit). Get event_id right.
Chapter 37. The server-side endpoint, batch limits, the two key types in Ads Manager, and the webhook-context pattern for payment webhooks.
Chapter 40. Match event_id between pixel and CAPI.

If you are building campaign management on top of the Advertiser API:

Chapter 41. Base URL, auth, rate limits, lifecycle states.
Chapter 42. Confirm your key works against GET /ad_account.
Chapters 43, 44, and 45. Create campaigns, ad groups, and ads.
Chapter 46 for programmatic asset upload, Chapter 47 for insights.

Debugging path. Each chapter is self-contained and source-cited. Open the chapter, find the field or endpoint you are debugging, cross-check against the source URL in the frontmatter. If the live OpenAI or Shopify doc has changed since last_verified, the upstream doc wins. This encyclopedia is a snapshot, not a fork.

The four pillars

Every chapter is organized under one of four pillars. Memorize them; they are how the discipline is structured.

Pillar	What it answers	Chapters
1. Technical Foundations	Can AI agents access your store?	4-7
2. Content Architecture	What do AI agents see when they read your store?	8-13
3. Off-Site Authority	What does the rest of the internet say about your brand?	14-16
4. Measurement & Iteration	How do you know any of it is working?	22-26

Channel-specific chapters (17-19) and the paid stack (25, 29-47) cut across all four pillars.

Conventions used throughout

Difficulty legend

The only emojis used in the encyclopedia.

🟢 Easy. No specialist required. Can be completed by a generalist marketer or store owner.
🟡 Medium. Needs marketing or developer skill.
🔴 Hard. Needs GEO-specific expertise.

These appear in the Difficulty column of every task table.

Cadence labels

The cadence column carries a text label (no emoji). Exact labels depend on chapter type:

Operational chapters. Real-time, Weekly, Monthly, Quarterly, Annual.
Definitional chapters. Pre-flight (foundational, one-time).
Technical chapters. Setup, Validation, Maintenance.
Channel-specific chapters. Onboarding, Activation, Optimization.
Strategic chapters. Taxonomy varies; the cadence column may be omitted.

Time estimates

Time estimates are not in chapter tables. They live in an internal master operations table, because:

Time-to-completion varies by store size, team, and tooling.
First-time implementers should multiply published estimates by 1.75x.
The encyclopedia is a reference document, not a project plan.

If you want time estimates per task, run an audit. The audit is calibrated to your store’s specific configuration.

“Last verified” stamp

Every chapter opens with a last_verified timestamp. AI commerce ships major updates monthly. A stamp older than 6 months means the chapter is a candidate for review. Stamps older than 12 months mean the chapter is actively decaying. Refer to the standalone post linked in the chapter’s “Deeper dive” section, or to the Sources & References master page, for any updates.

Source convention

Every chapter cites sources in three tiers:

Tier 1. Academic. Peer-reviewed papers and arXiv preprints (Aggarwal et al. 2024, Chen et al. 2025, AgenticGEO 2026). Cited for foundational claims, principles, and frameworks.
Tier 1. Official platform documentation. Shopify, OpenAI, Google, Perplexity, Microsoft. Cited for protocol, schema, API, and channel-specific facts.
Tier 2. Strong practitioner sources. Ahrefs, Adobe Analytics, Adthena, Seer Interactive, AirOps, Frase. Cited for empirical data and field-tested specifics where no academic or platform source exists.

The rule: cite the highest-tier source that supports the claim. If Aggarwal et al. proves it, do not cite a blog summary.

Citations are inline footnotes (e.g. [^1]). Each footnote contains a brief reference and a link to the back-of-book Sources & References page.

If a claim is not sourced, it is not in the encyclopedia. Hard rule.

A note on numbers (paid chapters)

The paid chapters (29-47) use two distinct numeric conventions. They look similar and are not the same.

Event monetary fields (pixel and Conversions API). Integers in the lowest currency denomination. 12999 means $129.99 in a USD account. Covered in Chapter 39.
Bid and budget fields (Advertiser API). Micros, millionths of the main currency unit. 60000 micros equals $0.06, or $60 CPM. Covered in Chapter 44.

Mixing the two is the most common Advertiser API integration bug.

What this encyclopedia does not cover

Off-Shopify ecommerce platforms. WooCommerce, Magento, BigCommerce, custom stacks. The frameworks transfer; the Shopify-specific tactics do not.
B2B ecommerce GEO. Different buyer behavior, different review surfaces, different sources cited by AI engines. Out of scope.
Branded SEO and demand generation. Referenced where relevant, but the encyclopedia assumes you already run paid acquisition and brand work.
Paid layer beyond ChatGPT Ads. Perplexity, Gemini, Claude, and Copilot do not currently offer generally available paid AI-search inventory. The encyclopedia adds chapters as those layers go live.