The official SEO guidelines live across six primary sources: Google Search Central, Google Business Profile help, the Bing Webmaster Guidelines, Schema.org, the robots.txt standard, and IndexNow. This page maps every one, tells you what each governs, flags the rules that catch people out, and links straight to the canonical doc. Bookmark it. We keep it current.
There is no single rulebook for SEO. The real rules are scattered across Google's, Microsoft's, and the open web's official documentation, and most "SEO guidelines" articles only summarise Google. This cheatsheet pulls all of it into one map: the primary sources, the specific pages worth bookmarking, and the gotchas the Hawk Academy course sees trip people up most. Every link goes to the source, not to an opinion about it.
Why we built this: at StudioHawk we settle arguments with the primary doc, not a blog post. This is the reference sheet our team and students actually use. For the practitioner version on a personal site, see Lawrence Hitches' Official SEO Guidelines Cheatsheet, this is the Hawk Academy teaching companion to it.
Google Search Central
Google Search Central is the official home of Google's SEO documentation. Start with Search Essentials (the rules formerly called the Webmaster Guidelines), then the SEO Starter Guide and the helpful-content docs. These three cover what Google requires, recommends, and rewards.
Search Essentials has three parts: technical requirements (the minimums to be eligible for Search), spam policies (what gets you demoted or removed), and key best practices (what actually moves rankings). The pages below are the ones worth keeping open.
| Doc | What it governs | Link |
|---|---|---|
| Search Essentials | The baseline: technical requirements, spam policies, and key best practices. Formerly the Webmaster Guidelines. | essentials |
| SEO Starter Guide | The official primer on how Search works and the basics that matter most. | seo-starter-guide |
| Creating helpful content | People-first content guidance, the self-assessment questions, and the YMYL framing. | creating-helpful-content |
| Title links | How Google generates the blue title in results and how to influence it. | title-link |
| Crawlable links | What counts as a followable link. If it is not an <a href>, Google may not follow it. | links-crawlable |
| Qualify outbound links | The rel="sponsored", rel="ugc" and rel="nofollow" attributes and when to use each. | qualify-outbound-links |
| Structured data gallery | Every structured-data type Google supports for rich results, with per-type docs. | search-gallery |
| AI features and your website | How Google's AI features (AI Overviews, AI Mode) use your content, and how to manage it. | ai-features |
The one students miss: "requirements" and "best practices" are not the same tier. The technical requirements are pass or fail for eligibility. Best practices are where rankings are won. Read our E-E-A-T guide and information gain guide for the two that matter most.
Google Business Profile and local
Local SEO has its own official rulebook in Google Business Profile help, separate from Search Central. The most-broken rule is the business description: it has a hard 750-character limit, and only the first 250 or so are shown before the fold.
| Doc | What it governs | Link |
|---|---|---|
| Business description | The 750-character limit, what is allowed, and what gets your profile flagged. | business/answer/3038177 |
Bing Webmaster Guidelines and AI Visibility
Microsoft publishes its own Bing Webmaster Guidelines covering discovery, indexing, evaluation and surfacing across Bing and Copilot. As of 2026, Bing Webmaster Tools also reports AI Visibility Insights, including Citation Share, Topics and Intents, the only first-party data on how an AI assistant uses your site.
This is the source almost nobody documents. Bing's AI Performance report shows your Citation Share (your slice of the sites an AI answer pulls from), the Topics it groups your content under, and the Intents behind the queries. It is observational, not a competitor scoreboard, but it is the closest thing to ground truth for AI visibility.
| Doc | What it governs | Link |
|---|---|---|
| Bing Webmaster Guidelines | Microsoft's official rules for discovery, indexing, evaluation and surfacing on Bing and Copilot. | webmasters-guidelines |
| AI Visibility Insights | The 2026 AI Performance report: Citation Share, Topics and Intents, inside Bing Webmaster Tools. | bing.com/webmasters |
First-mover note: Bing handed out AI citation data for free and almost nobody is reading it. Our full walkthrough is in the Bing Citation Share guide.
Schema.org and structured data
Schema.org is the shared vocabulary for structured data, maintained by Google, Microsoft, Yahoo and Yandex. Schema.org defines every type and property; Google's Rich Results Test tells you whether your markup is eligible for a Google rich result.
Two different jobs: Schema.org is the dictionary of what you can mark up, and Google's documentation is what Google will actually use. A type can exist on Schema.org and still produce no rich result in Google, so always check both.
| Doc | What it governs | Link |
|---|---|---|
| Schema.org type library | The full vocabulary of types and properties for structured data. | schema.org |
| Rich Results Test | Google's official tool to validate whether your markup is eligible for a rich result. | test/rich-results |
| FAQPage structured data | The FAQPage spec. Note the rich result was deprecated in 2026 (see quirks below). | structured-data/faqpage |
robots.txt and AI-crawler directives
The robots.txt standard is now a formal spec (RFC 9309), and the same file controls AI crawlers. To allow or block an AI bot, target it by user-agent token: GPTBot, ClaudeBot, PerplexityBot, or Google-Extended.
Blocking an AI crawler in robots.txt is the single highest-leverage AI-visibility decision most sites never make on purpose. Block GPTBot and you cannot be cited by ChatGPT. Google-Extended is the important nuance: it opts you out of Google's AI training and grounding without affecting normal Search indexing.
| Crawler | Operator | robots.txt token | Docs |
|---|---|---|---|
| GPTBot | OpenAI (ChatGPT) | GPTBot | openai bots |
| ClaudeBot | Anthropic (Claude) | ClaudeBot | anthropic crawler |
| PerplexityBot | Perplexity | PerplexityBot | perplexity bots |
| Google-Extended | Google (Gemini, AI training) | Google-Extended | google crawlers |
| Standard | What it governs | Link |
|---|---|---|
| robots.txt (RFC 9309) | The formal Robots Exclusion Protocol specification. | rfc9309 |
| MDN robots.txt | The plain-English reference for robots.txt syntax and behaviour. | MDN robots.txt |
| Google robots intro | How Google interprets robots.txt, with the edge cases. | robots/intro |
Check yours in seconds: our free AI Crawler Access Checker reads your robots.txt and tells you which AI bots you are blocking.
IndexNow
IndexNow is an open protocol for instantly telling search engines that a URL is new or changed, instead of waiting for a recrawl. Submit a URL to one participating engine and it is shared with all of them.
Participating engines in 2026 include Microsoft Bing, Yandex, Seznam and Naver. Google does not currently consume IndexNow, so it complements rather than replaces sitemaps and the Search Console URL tools.
| Doc | What it governs | Link |
|---|---|---|
| IndexNow | The protocol, the participating engines, and how to implement the ping. | indexnow.org |
Documentation quirks that catch people out
Some official rules contradict what SEOs still do out of habit. These are the ones that cause the most wasted effort and the most confident-but-wrong advice.
- Meta keywords are ignored. Google has not used the meta keywords tag for ranking in well over a decade. It is still in templates everywhere. Stop filling it in.
- FAQ rich results were deprecated in 2026. The visible FAQ snippet rarely shows now. Keep FAQ content and FAQPage schema for clarity and AI parsing, but do not expect the rich result.
- nofollow is a hint, not a directive. Since 2019 Google treats
nofollow,sponsoredandugcas hints it may choose to use, not hard rules. - Requirements are not best practices. Meeting the technical requirements makes you eligible. It does not make you rank. People conflate the two constantly.
- Crawl-delay is not a Google directive. Google ignores
crawl-delayin robots.txt. Use Search Console crawl-rate settings instead.
How to use this cheatsheet
Treat it as the index, not the encyclopedia. When a question comes up, find the source here, click through to the canonical doc, and quote the primary source rather than a summary. Three habits that pay off:
- Verify against the doc, not a blog. SEO advice ages badly. The official docs are the only thing that settles it.
- Re-check before you build. Rich-result support and AI directives change. Confirm the rule still holds before you ship.
- Map your AI access first. Before chasing AI citations, make sure you are not blocking the crawlers in robots.txt.
FAQ
Where can I find Google's official SEO guidelines?
In Google Search Central, specifically the Search Essentials page (formerly the Webmaster Guidelines), plus the SEO Starter Guide and the helpful-content documentation.
What is the official Bing equivalent of Google's guidelines?
The Bing Webmaster Guidelines, which cover discovery, indexing, evaluation and surfacing across Bing and Copilot. Bing Webmaster Tools also now reports AI Visibility Insights including Citation Share.
What is the difference between rel=nofollow, sponsored and ugc?
They are all link relationship hints. Use sponsored for paid or affiliate links, ugc for user-generated content like comments, and nofollow as a general "do not pass credit" signal. Google treats all three as hints.
Is FAQPage schema still worth adding in 2026?
The FAQ rich result was deprecated, so the snippet rarely appears. The content and the schema are still useful for clarity and for AI engines that parse the page, so keep them, just do not expect a rich result.
Which search engines support IndexNow?
Microsoft Bing, Yandex, Seznam, Naver and others. Submitting to one participating engine shares the ping with the rest. Google does not currently consume IndexNow.
How do I control AI crawlers like GPTBot and ClaudeBot?
Through robots.txt, by user-agent token: GPTBot for OpenAI, ClaudeBot for Anthropic, PerplexityBot for Perplexity, and Google-Extended to opt out of Google's AI use without affecting Search.
All sources, in one place
- Google: Search Essentials, SEO Starter Guide, Helpful content, Title links, Crawlable links, Qualify outbound links, Structured data gallery, AI features, Crawlers overview, robots intro.
- Google Business Profile: Business description.
- Bing: Webmaster Guidelines, Webmaster Tools.
- Structured data: Schema.org, Rich Results Test.
- robots and AI crawlers: RFC 9309, MDN, GPTBot, ClaudeBot, PerplexityBot.
- IndexNow: indexnow.org.
Spotted something out of date? This page is maintained, tell us and we will fix it. Last reviewed June 2026.