How long does GEO take to show results?

Three weeks for a 30 to 80 page site. Week one is the audit: existing schema, robots.txt, sitemap, AI crawler coverage. Week two is the technical fixes — identity graph in the HTML, FAQPage schema on top pages, BreadcrumbList, robots.txt rewrite. Week three is content shape: rewriting the top H2 sections so each opens with a self-contained answer. Beyond that, the work becomes ongoing content production.

Do I still need GEO if I rank well on Google?

Yes. AIO citation share and Google ranking position have low overlap — ranking #1 no longer reliably wins an AI Overview citation. You need to optimize for both. The work is largely the same, but the passage-level writing and entity schema required for AI citations often differ from what produced your Google rankings.

GEO FOR SMBS · 7 chapters

Generative Engine Optimization for SMBs, written by someone who ships it for a living.

By José Cabal · Last updated June 25, 2026

Generative Engine Optimization (GEO) is the practice of getting a website cited as a source inside AI-generated answers from ChatGPT, Perplexity, Claude, Copilot, and Google AI Overviews. This guide covers what GEO is, how each AI engine picks sources, the schema stack, crawler access, and the four-metric scorecard that proves it works.

Start with Foundations Get help implementing GEO

chapters covering foundations, per-platform tactics, schema and passages, crawler access, and measurement.

answer engines covered side by side — ChatGPT, Perplexity, Claude, Copilot, and Google AI Overviews + AI Mode.

fabricated statistics. Every external number links to a primary source from the vendor's own documentation.

Who this guide is for.

If one of these three is the week you are having, the guide will save you a quarter of trial and error.

Founder watching organic traffic flatten

Google still sends you sessions, but the volume that used to convert is being swallowed by AI Overviews and ChatGPT answers. You need to be the source the answer cites, not the link nobody clicks.

First marketing hire at a B2B SMB

Your CEO keeps asking why the company is not showing up when they ask ChatGPT about their own category. This is the guide that turns that into a fixable list, not a vague aspiration.

Agency or fractional CMO leading a relaunch

You are rebuilding a site and want GEO baked into the foundation — schema stack, passage discipline, crawler access — rather than retrofit it in year two.

The seven chapters.

Default reading order is top to bottom. Each chapter stands alone if you arrived from search or an AI citation — every page opens with a self-contained answer paragraph.

START HERE

GEO for SMBs: The Generative Engine Optimization Guide

A field guide for 5 to 50 person businesses that want to get cited inside ChatGPT, Perplexity, Claude, Copilot, and Google AI Overviews — not just rank in classical Google.

What Generative Engine Optimization Actually Is

GEO defined in plain English, how it differs from SEO and AEO, and why a 2026 SMB cannot treat AI search as a side experiment any more.

PLATFORMS

Getting Cited in Google AI Overviews and AI Mode

How Google's Gemini-grounded surfaces — AI Overviews inside the SERP and the standalone AI Mode — pick sources, and the page-level fixes that earn citations.

Optimizing for ChatGPT, Perplexity, Claude, and Copilot

What each major answer engine looks for, which retrieval bots they use, and the per-platform tweaks that move citation share without rewriting the whole site.

IMPLEMENT

Schema and Passage Writing for AI Citations

The structured data types that still matter post the May 2026 FAQ deprecation, and the passage-shape rules that make a paragraph quotable by a language model.

AI Crawlers, robots.txt, and llms.txt for SMBs

Which AI bots to allow, which to block, what to put in robots.txt and llms.txt, and how to keep training crawlers out without losing retrieval citations.

OPTIMIZE

Measuring AI Citations and GEO Performance

GA4 referral patterns, prompt-test panels, brand-mention monitoring, and the four-metric GEO scorecard that tells a founder whether the investment is working.

The five mistakes that quietly kill SMB citation share.

These are the most common failure patterns I see when auditing sites that say "we have done GEO" and have no citations to show for it.

Mistake	What it costs	The fix
Blocking all AI bots in robots.txt as a default.	Disqualifies the site from being cited inside ChatGPT, Perplexity, Claude, and Copilot answers — for many SMBs this already exceeds residual SEO traffic on commercial queries.	Allow retrieval bots (OAI-SearchBot, Claude-SearchBot, PerplexityBot, Perplexity-User, Bingbot). Decide on training bots (GPTBot, ClaudeBot, Google-Extended) separately on a licensing basis.
Writing for keyword density instead of passage extraction.	A page may rank in Google and never get cited by an AI engine because no paragraph stands alone.	Re-write the lead and each H2 to answer a specific question in 40 to 80 words. The page should still rank — and now it can be quoted.
Stuffing FAQPage schema on every page after Google's May 2026 deprecation.	Risks a spammy structured data manual action. Provides no SERP rich result. Confuses AI engines about what is actually a Q/A page.	Keep FAQPage only where genuine user-facing Q/A exists on the page. Remove it from product pages, service pages, and homepages that have no real Q/A.
No author or organization schema — pages float without an entity owner.	AI engines cannot attribute a citation cleanly, so they often pick another source instead.	Ship Person + Organization JSON-LD with sameAs links to LinkedIn, Wikidata, and any authoritative third-party profile. Reference both via @id from every Article.
Measuring GEO purely by GA4 referral sessions and concluding it does not work.	AI referrals are systematically under-counted in 2026 — many land in Direct, brand-search lifts get attributed to Organic, and zero-click answers leave no session at all.	Run a prompt panel monthly, configure an AI channel group in GA4, and track brand-mention share. Use referral sessions as one of four metrics, not the only one.

Common questions.

What is Generative Engine Optimization (GEO)?

Generative Engine Optimization is the discipline of making a website's pages get cited as sources inside answers from generative AI engines — ChatGPT, Perplexity, Claude, Microsoft Copilot, and Google's Gemini-grounded AI Overviews and AI Mode. GEO shares a technical foundation with SEO (indexability, schema, internal links) but optimises for citation share inside an AI answer rather than rank position on a SERP. Most SMBs treat it as a layer on top of existing SEO work, not a replacement.

How is GEO different from SEO?

SEO optimises for a ten-blue-link result; GEO optimises for being one of three to seven sources cited inside an AI-generated answer. The page-quality signals overlap heavily — index coverage, schema, internal links, content depth — but GEO additionally rewards self-contained passages, fact density, named-entity clarity, third-party presence (Wikipedia, Wikidata, LinkedIn), and an Author + Organization schema stack the model can attribute to.

Should an SMB block GPTBot, ClaudeBot, and Google-Extended?

Blocking the training crawlers (GPTBot, ClaudeBot, Google-Extended) keeps content out of future model training but does not affect whether the site is cited inside live answers. The retrieval bots (OAI-SearchBot, Claude-SearchBot, PerplexityBot, Bingbot for Copilot) are the ones that matter for citations. Most SMBs allow retrieval and decide on training case by case, based on the licensing posture they want to take.

Does Google AI Overviews still count as SEO or is it a separate discipline?

It is a separate surface that grounds on Google's regular index. The classical SEO foundation is required (indexed, schema, internal links) but AIO and AI Mode citations additionally favour passage-shaped content that answers the user's sub-queries cleanly. Treat AIO as the highest-leverage GEO surface for most SMBs because the existing Google work compounds into it.

How long does GEO take to show measurable results?

AI engines re-crawl and re-rank citations faster than Google ranks for new content — meaningful citation share shifts can show up inside 30 to 90 days on focused queries. The slow part is building the entity graph (Wikidata, LinkedIn, sameAs across third-party profiles) that AI engines weight heavily; that part is a quarters-not-weeks investment.

Why is this guide written by José Cabal?

José is a Vancouver AI marketing consultant who has implemented GEO work for B2B SMBs since AI Overviews first rolled out. The guide reflects field experience: what moves citation share, what is theatre, and where small businesses get the most ROI out of a limited content budget. Every claim links to the engine vendor's own documentation so it can be re-verified.

COMPANION GUIDES

GEO is the visibility layer. The system around it matters too.

The AI Marketing Playbook is the broader operating system GEO lives inside, the AI Lead Generation guide is the bottom-of-funnel layer that turns AI-driven visibility into pipeline, and the HubSpot for SMBs guide is where the data and reporting actually live.

Read alongside the guide

Field reports and decision essays adjacent to the chapters above.