SEO vs AEO vs GEO vs MEO: How Each Kind of Search Works

By Garin Froehlich, Founder, GMScalezLast updated: November 202618 min read

A customer can now find a local business in four places: the classic list of results on Google, the answer box or AI summary above that list, an AI assistant such as ChatGPT, and the map. Each is sorted by a different method and rewards different things. SEO (search engine optimization) earns a spot in the classic list, AEO (answer engine optimization) earns a place in answers, GEO (generative engine optimization) earns a mention in AI assistants, and MEO (map engine optimization) earns a place in map results. All four rest on one foundation: clear facts about the business, pages that each answer one question, structured data that matches the page, real reviews, and mentions on other sites.

Key takeaways

  • Classic search crawls, indexes and ranks, and Google names the ranking systems that do the sorting, including link analysis, which still counts.
  • AI Overviews are assembled from many related searches, and most cited passages come from the top of a page.
  • AI assistants use their own crawlers, and the major ones measured do not run JavaScript, so a page must serve its full content in the raw code.
  • Map results are mainly based on relevance, distance and popularity, and practitioners rank the primary profile category first.
  • Brand mentions across the web track AI Overview visibility more closely than backlinks do, though that is a correlation, not proof of cause.
  • Each kind of search needs its own measurement: Search Console, the Analytics "AI Assistant" channel, map-grid checks and profile actions.
  • No one can buy a better local ranking, and no special file replaces the shared foundation.

Four kinds of search, one business

The four acronyms name four places where a person can meet a business, and each place sorts candidates its own way. The table defines each term.

TermSEO
What it isEarning a position in the ordinary list of results
Where the customer sees itThe blue-link results on Google and Bing
What it rewards mostPages that answer the query, links from other sites, results searchers choose
TermAEO
What it isEarning a place in answer boxes and AI summaries
Where the customer sees itAn answer or AI Overview above the results
What it rewards mostSelf-contained answers near the top of a page, content that adds something new
TermGEO
What it isEarning a mention in answers written by AI assistants
Where the customer sees itChatGPT, Gemini and similar assistants
What it rewards mostPages a crawler can read in raw code, sourced facts, mentions across the web
TermMEO
What it isEarning a place in map results
Where the customer sees itThe map and the local listings beneath it
What it rewards mostThe right primary category, closeness to the searcher, a strong profile with real reviews

These are not four jobs. A mobile mechanic who is clearly described on the web and well reviewed by real customers has done most of the work for all four at once. The differences matter because each surface has its own measurements, and knowing them shows where effort goes furthest.

SEO: how a classic search engine decides what ranks

A classic search engine finds pages by crawling, files them in an index, and ranks the indexed pages against each query using scoring systems. Crawling means a program follows links to discover pages. Indexing means the engine reads a page and stores it for retrieval. Ranking is the sorting that happens when someone types a query.

The named ranking systems

Google does not rank with one formula. In "A guide to Google Search ranking systems", Google names its systems, including BERT, neural matching, RankBrain, passage ranking, freshness, link analysis and PageRank, the reviews system, original content, site diversity, and SpamBrain, its spam detection. It adds that the helpful content system was folded into the core ranking systems in March 2024.

Two details matter. Google says PageRank "continues to be part of our core ranking systems," so links from other sites still count. And it says: "Having some good site-wide signals does not mean that all content from a site will always rank highly." Each page has to earn its own position, which is why a dedicated page for each service matters.

Links and click behavior

PageRank treats a link from one page to another as a vote of confidence, with votes from trusted pages weighing more. Whitespark's Local Search Ranking Factors 2026, a survey of local search practitioners, ranks a dedicated page for each service first among local organic factors, geographic keyword relevance of content second, quality of inbound links third, and internal linking across the site sixth.

SparkToro's 2024 report quotes a Google vice president of search, cross-examined in the US antitrust trial, on a system called Navboost that dates to about 2005 or earlier, and says leaked Google documents tie Navboost to click data. The sensible reading is to write titles and descriptions for people, because a result people choose is behaving the way a click-aware system would reward.

AEO: how answer boxes and AI Overviews are built

An AI Overview is assembled from passages of pages that Google judges to answer the question directly, found through several related searches and not just one. A page can therefore be cited without holding the top organic position.

Query fan-out

Google's "AI features and your website" says AI Overviews and AI Mode "may use a 'query fan-out' technique" that issues multiple related searches across subtopics and data sources. The same page says AI Overviews appear only when Google's systems judge them additive to classic Search, and that there are no additional requirements or special optimizations to appear.

In practice, one question such as how a mobile mechanic replaces a starter splits into smaller ones about timing, tools and service area. A site that covers those subtopics on separate, focused pages gives the fan-out more to retrieve.

Why the first lines of a page matter

A 2026 CXL study of 100 AI Overview citations found that 55% of the cited passages came from the first 30% of the page. The answer belongs at the top, written so it still makes sense when lifted out and read alone.

Google's May 2025 post, "Top ways to ensure your content performs well in Google's AI experiences on Search", points the same way: unique, non-commodity content, good page experience, crawl access, and structured data that matches the visible content. Structured data is code that labels a page's facts, such as name, hours and services, in a format machines read reliably.

What happens to clicks

Pew Research Center's July 2025 study covered 68,879 searches by 900 US adults. People clicked a traditional result on 8% of visits when an AI summary appeared and 15% when it did not, and clicked a link inside the summary on 1% of visits. Ahrefs found that when an AI Overview appears, the click-through rate of the top-ranking page was about 34.5% lower in its first analysis and about 58% lower in its December 2025 data.

SourcePew (July 2025)
FindingTraditional result clicked on 8% of visits with an AI summary, 15% without
SourcePew (July 2025)
FindingLink inside the summary clicked on 1% of visits
SourceAhrefs
FindingTop page's click-through rate about 34.5% lower at first, about 58% lower in December 2025 data
SourceCXL (2026)
Finding55% of cited passages from the first 30% of the page

Being named in the answer is now part of being found, even when a click does not follow.

GEO: how AI assistants find and cite sources

AI assistants gather pages with their own crawlers, read only what those crawlers can see, and write answers that name the sources and brands they trust. Winning here means being readable to the crawler and mentioned widely.

Their crawlers

OpenAI's crawler documentation names three. OAI-SearchBot gathers pages for ChatGPT search, GPTBot gathers pages for model training, and ChatGPT-User fetches a page when a user asks. They are controlled separately in robots.txt, the file that tells crawlers where they may go, so blocking GPTBot does not remove a site from ChatGPT search.

Why raw page code matters

Many sites build their content in the visitor's browser using JavaScript, a language that runs on the page. Vercel and MERJ reported in December 2024 that the major AI crawlers they measured did not run JavaScript, so they read only the raw page code. A page that looks complete to a person can look empty to such a crawler. Every page should serve its headings, services and contact details before any script runs.

What the research says raises visibility

The paper "GEO: Generative Engine Optimization" (Aggarwal et al., KDD 2024) found that adding statistics, quotations and cited sources raised a page's visibility in generative answers by up to 40%, while keyword stuffing offered little to no improvement. Assistants favor pages that read like a reliable reference.

Why brand mentions matter more than links

In a 2025 study of 75,000 brands, Ahrefs found that how often a brand is mentioned across the web correlated 0.664 with its visibility in AI Overviews, against 0.218 for backlinks. That is a correlation, not proof of cause, but it suggests being consistently named and described on many sites matters.

Where the mentions live matters too. Steady Demand ran 2,880 prompts across 12 US metros in August 2026. Yelp data was attached to ChatGPT's local business cards in 95.83% of runs, Google data in 4.62%, and Foursquare appeared in none of the layers measured. A business card is the boxed listing ChatGPT shows for a local business, with its details drawn from outside sources.

Our method behind those numbers: each service has its own page, every page serves its full content in the raw code, and structured data on every site covers the business, services, areas served, hours, rating and real reviews. Every client domain is verified in Bing Webmaster Tools and pushes changes through IndexNow, a protocol that tells participating search engines a page has changed. The business's profiles across the web are tied together in its structured data.

MEO: how map results are ranked

Map results are ranked mainly on relevance, distance and popularity. Google's "Tips to improve your local ranking on Google" says: "Local results are mainly based on relevance, distance, and popularity." Relevance is how well a profile matches the search, distance is how far the business is from the searcher, and popularity, also called prominence, is how well known the business is.

Google explains that prominence is based partly on how many websites link to the business and how many reviews it has, and says: "There's no way to request or pay for a better local ranking on Google."

The factors practitioners rank highest

The Whitespark 2026 survey ranks these factors highest for the local pack, the group of map listings shown within search results.

Rank1
Local pack factorPrimary Google Business Profile category
Rank2
Local pack factorProximity of the address to the point of search
Rank3
Local pack factorKeywords in the business title
Rank4
Local pack factorPhysical address in the city of search
Rank5
Local pack factorBusiness open at the time of search
Rank6
Local pack factorHigh Google ratings
Rank8
Local pack factorAdditional categories
Rank9
Local pack factorNumber of Google reviews with text
Rank11
Local pack factorRecency of reviews
Rank13
Local pack factorClick-through rate from map results
Rank14
Local pack factorA sustained flow of reviews

The profile's own fields, category, title, hours and address, fill most of the top of the list, and those are within the owner's control. For a mobile business, distance is measured from the verified location, which is why we check positions from a grid around it and not from one point.

What all four share

All four depend on five things: clear facts about the business, pages that answer one question each, structured data that matches the page, reviews, and mentions on other sites.

  • Clear facts.

    Name, address, phone, hours, services and service area should match everywhere they appear, because search, maps and assistants cross-check them.

  • One question per page.

    A dedicated page per service ranks first among local organic factors, and a focused page gives answer boxes and assistants a clean passage to lift.

  • Matching structured data.

    Markup should describe the business, services, areas served, hours and rating exactly as the visible page does, with real reviews only.

  • Reviews.

    They feed the map factors, Google's prominence explanation, and the structured data that reports ratings.

  • Mentions elsewhere.

    Mentions correlate with AI visibility, links feed prominence, and listings such as Yelp ground local answers.

How to measure each one

Each kind of search leaves a different trail, so each needs its own measurement.

Kind of searchSEO
Where to lookSearch Console, reviewed weekly
What it showsClicks, queries and pages for classic results
Kind of searchAEO
Where to lookSearch Console query data, plus checks of AI Overviews
What it showsQuestion-style searches arriving and whether the business is named
Kind of searchGEO
Where to lookThe Analytics "AI Assistant" channel
What it showsAssistant-referred sessions, engagement and key events
Kind of searchMEO
Where to lookMap-grid checks and profile actions
What it showsPosition across the area, and calls, directions and website clicks

Google Analytics Help lists an "AI Assistant" channel in its default channel groups that collects visits referred by assistants such as ChatGPT and Gemini, and counts visits from Google's AI Overviews and AI Mode under Organic Search. Search Engine Journal reported on May 14, 2026 that Google Analytics had added the channel.

A map grid is a set of points laid over the service area, with the ranking recorded at each. Ours is nine points, two miles apart, centred on the profile's verified location, checked weekly. Analytics key events show which pages turn visits into calls and texts, and we make ranking claims only on city-specific searches.

Myths worth dropping

  • A special file such as llms.txt gets you into assistants.

    Search Engine Roundtable quoted Google's John Mueller: "no AI system currently uses llms.txt".

  • Blocking GPTBot removes you from ChatGPT.

    OpenAI's documentation says the crawlers are controlled separately.

  • AI Overviews need a special optimization.

    Google says there are no additional requirements.

  • More keyword repetition wins.

    The GEO research found keyword stuffing offered little to no improvement.

  • Backlinks are the whole story.

    PageRank is still part of Google's core ranking systems, yet mentions correlated more strongly with AI Overview visibility in the Ahrefs study.

  • Local ranking can be bought.

    Google says there is no way to pay for it.

  • Foursquare feeds ChatGPT's local answers.

    In the Steady Demand runs it appeared in none of the layers measured, while Yelp data was attached to business cards in 95.83% of runs.

What search data tells you about running the business

All four kinds of search leave a record of what customers want, and that record is useful well beyond marketing. Search Console shows which services people look for and in which towns. Google Business Profile reports show how many people call, ask for directions or visit the website, and Analytics shows which pages lead to a call or a text. Read together, they show which services are in demand, where the calls come from and at what time of day.

Owners can use that to decide what to charge for the work people ask about most, which hours to keep, which towns belong in the service area, which services earn a place on the menu and when a second technician makes sense. Working with us includes this kind of business advice. Each month we look at the search demand, the call volume and the competitors in each market, and we talk through pricing, hiring and the service mix with the owner as part of the work. Our business growth advice page explains how that works, and pricing your services is a good place to start.

How to apply this

In priority order for a local service business:

  1. Set the right primary Google Business Profile category,then add the additional ones that truly fit.

  2. Make business facts identical everywhere:website, profile and every directory, including Yelp.

  3. Give every service its own page,with the places you serve on pages that name them.

  4. Open each priority page with a self-contained answerin the first 100 words, and follow question-style headings with a 40 to 60 word direct answer.

  5. Serve the full page in the raw code,and check that content appears before any JavaScript runs.

  6. Add structured data that matches the page,covering the business, services, areas served, hours, rating and real reviews.

  7. Let the crawlers in.Allow OAI-SearchBot in robots.txt, verify the domain in Bing Webmaster Tools, and use IndexNow.

  8. Keep the profile active and true:posts, photos, a services list that matches the site, accurate hours, and a reply to every review.

  9. Earn reviews and mentions steadily,and add pages for real search gaps found in Search Console.

  10. Measure each surface on its own terms,as in the table above.

FAQ

AEO sits on top of SEO. Google says AI features need no special optimization beyond what supports classic Search. The difference is in the writing: a self-contained answer at the top of each page.

No. The same pages serve both when the full content is in the raw code and each page answers its question clearly. Only the crawler and the measurement differ.

Yes. Ranking still decides who gets cited, and being named in an answer builds recognition even when no click follows. Pew found a traditional result clicked on 15% of visits without a summary.

Look for the "AI Assistant" channel in Google Analytics. Visits from Google's AI Overviews are counted under Organic Search, not in that channel.

Google measures from the business's verified location, so a traveling business is still ranked by closeness to it. Category, reviews and service pages then carry more of the weight.

Blocking GPTBot stops training use only, while OAI-SearchBot decides whether pages can appear in ChatGPT search. A service business that wants to be found usually allows the search crawler.

Yes. Search and profile reports show which services customers want, which towns they call from and when, and that informs pricing, hours, service area and hiring. We cover this kind of advice as part of our work, and the business model and service area page shows how we approach it.

Closing

We apply these principles for the mobile and service-area businesses we work with, building each plan around the business in front of us. Every client receives 8 off-site articles and 2 on-site articles of 800 words or more each month, each aimed at a real search gap found in Search Console. We review Search Console history weekly, and Analytics key events show which pages turn visits into calls and texts. From that data we adjust the work month by month. The conversation starts with a free one on one marketing consultation.

Related guides and pages

Call or text us at (712) 251-1967. Email us at garinfroehlich37@gmail.com.

Last updated: November 2026

CallTextFree consultation