Lawrence Hitches Written by Lawrence Hitches | AI SEO Consultant | May 05, 2026 | 8 min read

This is a quick-reference cheatsheet to the official SEO guidelines that matter most in 2026: Google Search Central, Bing Webmaster, Schema.org, MDN's robots.txt spec, and IndexNow. Each entry summarises what the official documentation says, what it actually means in practice, and links to the source. I built this because the official docs are scattered across half a dozen sites and dozens of pages. When you're trying to confirm a single rule (e.g. "does Google still support FAQ schema on non-FAQ pages?") it's faster to scan one page than to hunt through five.

Google Search Central, The Core Pages You Should Bookmark

Title Link Best Practices

Google's official guidance on writing title elements (the source for what becomes the title link in search results). Keep titles unique, descriptive, and concise. Google may rewrite your title link if it finds a better match for the query, typical rewrite rate is 40-60% across the web.

In practice: write titles 50-60 characters, lead with the primary keyword, include brand at the end on supporting pages but at the start on commercial money pages.

developers.google.com/search/docs/appearance/title-link

Heading Tags (SEO Starter Guide)

Use heading tags to signal structure. The starter guide doesn't say "you must have one H1" but it does say to use headings to convey structure and meaning. Skipping heading levels is allowed but discouraged. Heading text matters more than the level number for ranking signal.

In practice: one H1 per page (covered in our header tags guide), H2s for main sections, H3s only when an H2 has multiple subsections.

developers.google.com/search/docs/fundamentals/seo-starter-guide

Qualify Outbound Links: Sponsored, UGC, Nofollow

Three link attributes to declare relationships:

  • rel="sponsored", paid, affiliate, advertising links
  • rel="ugc", user-generated content (comments, forum posts)
  • rel="nofollow", generic "don't pass equity"; still the default if you can't be more specific

In practice: always disclose paid links with sponsored. Default to nofollow for generic untrusted links. Use ugc on comment sections.

developers.google.com/search/docs/essentials/spam-policies

Structured Data: FAQPage Guidelines

FAQPage schema is supported but with restrictions since 2023: Google only displays FAQ rich results from "well-known authoritative" government and health sites. Other sites can still mark up FAQs (and AI search engines like Perplexity and ChatGPT use the schema as an extraction signal), but the visible SERP enhancement is gone for most domains.

In practice: still ship FAQPage schema. The SERP rich result is gone for most sites, but AI engines parse it. We auto-generate it on every article on this site.

developers.google.com/search/docs/appearance/structured-data/faqpage

Helpful, Reliable, People-First Content (incl. YMYL)

Google's content quality framework. The "Your Money or Your Life" (YMYL) categories, finance, health, safety, legal, are held to higher E-E-A-T standards. The 2022 Helpful Content Update made content quality a sitewide signal: a domain with a lot of unhelpful content gets demoted across the board.

In practice: for YMYL topics, name the author (with credentials), cite sources, show direct experience. For everything else, write content that genuinely helps a human reader more than the next-best-ranking competitor.

developers.google.com/search/docs/fundamentals/creating-helpful-content

AI Features and Your Website

Google's documentation on how AI Overviews and Search Generative Experience use content. Standard SEO best practices apply, there's no separate "AI optimisation" requirement. Google does provide opt-out controls via Google-Extended in robots.txt for excluding content from training (covered in our robots.txt guide).

developers.google.com/search/docs/appearance/ai-features

Google Business Profile, Local SEO Documentation

Business Description: 750 Character Limit

The "from the business" section in Google Business Profile is capped at 750 characters. Google's official guidelines say: don't use it for promotional content, sales, or links. Use it to describe what the business does and what makes it distinct.

In practice: first 250 characters carry the most weight (visible without "read more"). Lead with what you do + city + the differentiator that matters most.

support.google.com/business/answer/7091

Bing Webmaster Guidelines, Often Forgotten, Worth Reading

Bing Webmaster Guidelines for AI Search

Bing officially documents how its AI search (powering Microsoft Copilot, ChatGPT search, and the broader OpenAI ecosystem) uses website content. Bing recommends: clear, well-structured content; semantic HTML; structured data; and accessibility best practices. Bing also officially supports IndexNow for instant indexing.

In practice: Bing has been undervalued by SEOs for a decade, but in 2026 it powers ChatGPT search, Copilot, and an increasing share of AI-mediated queries. Worth optimising for. See our Google vs Bing guide for the case.

bing.com/webmasters/help/webmasters-guidelines

Schema.org, The Structured Data Reference

Schema.org Type Library

Schema.org is the canonical reference for structured data types. The hierarchy starts at Thing and branches into hundreds of specialised types: Article, Product, LocalBusiness, Person, Organization, Event, etc.

Specialised LocalBusiness sub-types worth knowing (Google supports them all):

  • Restaurant, for restaurants, cafes, bars
  • MedicalBusiness, for clinics, doctors, hospitals
  • HealthAndBeautyBusiness, for spas, salons, beauty clinics
  • AutomotiveBusiness, for dealerships, mechanics
  • LegalService, for law firms
  • FinancialService, for accountants, financial advisors

In practice: use the most specific type that applies. HealthAndBeautyBusiness beats generic LocalBusiness for a salon, Google reads it as a stronger entity signal.

schema.org/docs/full.html

Structured Data Testing

The official validators:

Robots.txt, MDN and the Robots Exclusion Standard

MDN Robots.txt Documentation

Mozilla Developer Network's robots.txt page covers the core spec: User-agent, Disallow, Allow, Crawl-delay (Google ignores this), Sitemap. The 2022 RFC 9309 formalised the de-facto standard most major search engines have followed for two decades.

In practice: our robots.txt guide covers the 2026 reality including AI crawler controls (GPTBot, ClaudeBot, PerplexityBot, etc.) which aren't in the original spec but are now de-facto standard.

developer.mozilla.org/en-US/docs/Glossary/Robots.txt

IndexNow, Faster Indexing Across Search Engines

IndexNow Participating Search Engines (2026)

IndexNow is an open protocol that lets sites push instant indexing pings to multiple search engines at once. As of 2026, the participating engines are:

  • Bing (Microsoft), original co-founder of the protocol
  • Yandex, co-founder; matters for sites targeting Russian-language audiences
  • Naver, Korea's dominant search engine; added 2023
  • Seznam, Czech Republic; added 2022
  • Yep, privacy-focused search; added 2023
  • DuckDuckGo, partial integration via Bing's index

Notably absent: Google still doesn't participate. Google's ecosystem uses its own Indexing API (limited to JobPosting and BroadcastEvent schema) and the standard sitemap + crawl flow.

In practice: IndexNow integration is cheap (one HTTP POST per URL change) and gets your content into Bing, and through Bing, ChatGPT search and Copilot, in minutes instead of days. Worth wiring up if your site has frequently-updated content.

indexnow.org/searchengines

Google's Documentation Quirks That Catch People Out

A few things the official docs say that contradict common SEO advice:

  • Meta keywords aren't used. Confirmed by Google in 2009 and reaffirmed every couple of years. Stop adding them.
  • Word count isn't a ranking factor. Long content can rank because it covers more relevant subtopics, not because of length itself.
  • The disavow tool is rarely needed. Google's own docs since 2020 explicitly say most webmasters shouldn't need it. Use only if you have a manual action.
  • Multiple H1s are technically allowed in HTML5. The starter guide doesn't say "only one H1", that's SEO folklore. (Practitioner advice: still use one. More on this.)
  • JavaScript-rendered content gets indexed. Google renders JS; the older "JS = invisible to Google" advice is wrong as of 2018. There are still performance and discoverability trade-offs, but indexability isn't one of them.

How to Use This Cheatsheet

Three workflows:

  1. Quick fact-check. Want to confirm what the official docs say about a specific topic? Use this page as a jump-off to the source.
  2. Stakeholder evidence. When a client (or boss) pushes back on SEO advice, link them to the official source. Authority by reference.
  3. AI search context. AI engines (ChatGPT, Perplexity, Claude) cite official documentation heavily. Building a clean reference page like this gets cited because it consolidates information across multiple authoritative sources.

FAQ: Official SEO Guidelines

Where can I find Google's official SEO guidelines?

Google's main hub is Google Search Central (developers.google.com/search). The most-referenced pages are the SEO Starter Guide, the Search Essentials, the Spam Policies, and the structured data documentation.

Does Bing have an equivalent to Google Search Central?

Yes, Bing Webmaster Guidelines. Less polished than Google's documentation, but worth reading because Bing now powers ChatGPT search, Copilot, and a growing share of AI-mediated traffic.

Are Google Search Central guidelines requirements or recommendations?

Mostly recommendations. The Spam Policies are enforceable rules (violations can trigger manual actions). The other documentation is best-practice guidance, sites can rank without following every recommendation, but following them increases the probability of strong organic performance.

Does Google really not use meta keywords?

Correct, Google has not used the meta keywords tag for ranking since 2009. Officially confirmed and reaffirmed multiple times since. Bing and other engines also ignore it. Stop adding meta keywords to your pages.

What's the difference between rel="nofollow", rel="sponsored", and rel="ugc"?

All three are link attributes that tell search engines not to pass full ranking equity. rel="sponsored" declares paid or affiliate relationships. rel="ugc" identifies user-generated content. rel="nofollow" is the generic catch-all. Google introduced the more specific tags in 2019 to encourage better classification, they all function similarly but provide more accurate signal.

Is FAQPage schema still worth using?

Yes, but for different reasons than in 2021. Google restricted FAQ rich results to government and health authority sites in 2023, most sites no longer get the visible SERP enhancement. However, AI search engines (ChatGPT, Perplexity, Claude, Google AI Overviews) parse FAQPage schema as a strong extraction signal. Mark up FAQs to feed AI citations, not for the SERP rich snippet.

Which search engines participate in IndexNow?

As of 2026: Bing, Yandex, Naver, Seznam, Yep, and (partially) DuckDuckGo. Google does not participate, they use their own Indexing API for limited content types and standard crawl for everything else. Implementing IndexNow gets your content into Bing's index (and through Bing, into ChatGPT search and Copilot) in minutes instead of days.

Sources & Further Reading

Keep Reading

Soaring Above Search

Weekly AI search insights from the front line. One newsletter. Six sections. Everything that actually moved this week, with a practitioner's take.

Lawrence Hitches
Lawrence Hitches AI SEO Consultant, Melbourne

Chief of Staff at StudioHawk, Australia's largest dedicated SEO agency. Specialising in AI search visibility, technical SEO, and organic growth strategy. Leading a team of 120+ across Melbourne, Sydney, London, and the US. Book a free consultation →