---
title: "The AI Search Glossary for Owners Being Pitched AEO"
description: "A vendor says you have 38% share of voice in ChatGPT. Before you celebrate, ask whether that means a third of your market or appearances in 40 prompts…"
url: "https://groas.com/post/the-ai-search-glossary-for-owners-being"
image: "https://groas.com/media/blog/49541cca7630a6a1b6238805ff04cb139b04352c33b943c873f05d03b6a04f72.png"
published: "2026-10-06T05:23:56.840Z"
modified: "2026-10-06T05:23:57.870Z"
---

[AI For Google Ads](https://groas.com/category/ai-for-google-ads) · October 6, 2026 · 11 min read

# The AI Search Glossary for Owners Being Pitched AEO

[DavidFounder & CEO @ groas](https://groas.com/author/david)

![Cartoon of a vendor pitching wildly while a business owner secretly reads a 'Vendor Phrasebook' under the table, treating sales jargon as a foreign language.](https://groas.com/media/blog/49541cca7630a6a1b6238805ff04cb139b04352c33b943c873f05d03b6a04f72.png)

In this article

1. [Mention vs citation vs recommendation](#mention-vs-citation-vs-recommendation)
2. [Share of voice in AI search](#share-of-voice-in-ai-search)
3. [Ranking in an AI answer](#ranking-in-an-ai-answer)
4. [AI Overview vs AI Mode vs ChatGPT search](#ai-overview-vs-ai-mode-vs-chatgpt-search)
5. [AI Overview ads](#ai-overview-ads)
6. [Answer-first content](#answer-first-content)
7. [Passage](#passage)
8. [Entity and knowledge graph](#entity-and-knowledge-graph)
9. [Content gap (as AI sees it)](#content-gap-as-ai-sees-it)
10. [Verified bot vs spoofed User-Agent](#verified-bot-vs-spoofed-user-agent)
11. [Live-answer fetch](#live-answer-fetch)
12. [Rendering and blocked resources](#rendering-and-blocked-resources)
13. [robots.txt and AI user agents](#robotstxt-and-ai-user-agents)
14. [llms.txt](#llmstxt)
15. [Prompt tracking](#prompt-tracking)
16. [Tracked prompt caps](#tracked-prompt-caps)
17. [Ranking](#ranking)

A vendor says you have 38% share of voice in ChatGPT. Before you celebrate, ask whether that means a third of your market or appearances in 40 prompts the vendor chose.

Most AEO jargon is old PPC and SEO vocabulary with the meaning quietly changed. Once the answer is a paragraph, **a mention is not a citation**, an AI Overview ad is not an endorsement, and a rising score may say more about the prompts than your business. I spent years buying clicks when an impression and a ranking had fairly clear meanings. Here is how I read those words now, and what I ask when a deck uses them.

## Mention vs citation vs recommendation

**A mention names you, a citation links you, and a recommendation picks you.** [A mention puts your brand in the answer text without a link](http://ziptie.ai/blog/citation-rate-vs-mention-rate/). A citation attaches your URL as evidence for a claim. A recommendation tells the buyer you are the option to choose.

Decks often roll all three into “visibility.” That hides the awkward case: your page supplies the evidence while a competitor gets the recommendation. Researchers call this the _Mention-Source Divide_. I used to tell clients any visibility was good visibility. I was wrong. Being the footnote is not the same as getting the call.

Ask the vendor for three separate counts on the same prompts: **named, linked, recommended**. If those disappear into one score, you cannot tell what you bought.

## Share of voice in AI search

**Share of voice in AI search is the percentage of tracked answers where your brand appears.** The important word is _tracked_. The possible questions buyers can ask are effectively endless; the vendor chooses the sample. A report showing appearances in 15 of 40 prompts describes those 40 prompts, not a third of your market.

The score can move even when your site does not. [After OpenAI moved to ChatGPT 5.0, outbound citations fell across the platform and vendor dashboards moved with them](https://searchengineland.com/ai-share-of-voice-metrics-that-matter-more-479611). If an engine links out less often, a citation-based score can fall without a loss of brand relevance.

Use share of voice as a directional measure across the **same prompts and engines**. Ask to see the prompt list and the appearances split into mentions, citations, and recommendations. Without the denominator, the percentage is decoration.

## Ranking in an AI answer

**An AI answer does not give your business one numbered rank.** It may name a brand in its verdict, link several sources, and leave other businesses out entirely. Classic position two can still send traffic from the results page, but it does not guarantee a place in the answer above it.

I kept rank in reports after it stopped describing what I needed to know. Clients liked seeing a number rise. It made meetings easier and decisions worse. [An LLM response may cite only two to seven domains](http://ziptie.ai/blog/citation-rate-vs-mention-rate/); being one of those links still differs from being the company the sentence recommends.

For each prompt, ask: **Were we named, linked, or recommended, and where?** An average position cannot answer that.

## AI Overview vs AI Mode vs ChatGPT search

**AI Overview, AI Mode, and ChatGPT search are different places for a buyer to encounter an answer.** [An AI Overview appears within Google’s classic results page](https://avidopenaccess.org/using-ai-mode-within-google-search/), where ordinary listings remain nearby. AI Mode is a separate conversational experience. ChatGPT search and Perplexity sit outside Google and have their own citation habits.

The deck trick is to show one screenshot and report one score across every surface. That erases a practical difference: a classic listing can still earn a click alongside an Overview. In AI Mode, there is no classic listing alongside it. Ask the vendor to **split results by surface** before deciding whether a change in visibility matters.

## AI Overview ads

**An AI Overview ad is paid placement, not a citation.** Your brand can appear near an AI answer because an ad was eligible to show there, not because the answer chose your page as evidence. Stop paying for the placement and the ad goes away.

That sounds obvious until a report puts paid appearances and citations on the same slide. If you spend $20k a month and see your brand beside an Overview, you need to know whether the appearance came from your ad spend or from the answer linking your content. [Ads can appear above, below, or within an AI Overview](https://searchengineland.com/google-ai-overviews-where-ads-appear-456211). Those placements are still ads.

Ask the vendor to **label every appearance paid or cited**. Do not pay a separate citation fee for your own ad doing its job.

## Answer-first content

**Answer-first content opens with a direct, self-contained answer before it expands.** A summary box pasted onto a vague article does not make the article useful to an answer engine. The passage has to make sense when lifted out of the page: what is the answer, and what on the page supports it?

The familiar failure is 800 words of windup before the useful sentence. A model looking for a quotable answer can use the competitor’s clear passage instead. Ask the vendor: **Which exact passage do you want cited, and what did you cut to make it stand alone?**

![Ink cartoon of a vendor pointing to a footnote while a competitor receives the customer trophy](https://groas.com/media/blog/3fd3820d214ec2d5d8cc84a16427059fe7229ddca577c7fdb32534ee29983df1.png)

## Passage

**A passage is the chunk of a page an answer engine retrieves or cites.** The distinction matters because decks sell page-level fixes while the useful answer may live, or fail to live, in one paragraph of a long guide.

I learned to look for the block doing the work in PPC landing-page tests. The same habit helps here: find the sentence a model can quote without rewriting it. Then check whether a bot can read it. [Many AI crawlers read raw HTML without executing JavaScript](https://searchengineland.com/guide/ai-crawlers), so a passage injected after the page loads may be invisible to them.

Ask for the proposed **cited passage in the page’s HTML**, not a screenshot of the finished design.

## Entity and knowledge graph

**An entity is a distinct business or thing, with attributes and relationships a search system can recognize.** For a business, that means details such as its name, location, services, and connections to other sources, not just the keywords on a page. [Entity work aims to reduce ambiguity in Google’s Knowledge Graph](https://www.searchenginejournal.com/entities-foundation-modern-seo/562755/).

The misuse is prescribing ten blog posts when the business information itself conflicts across the site, Google Business Profile, and third-party listings. More prose cannot neatly resolve a split identity. I used to reach for content when a citation was missing. Sometimes the page was fine; the business record needed the work.

Ask: **What information about us conflicts, and where?** Fix that before buying another stack of posts.

## Content gap (as AI sees it)

**A content gap is a buyer question your competitors answer in a usable passage and you do not.** It might be a comparison, a price range, or an objection the answer needs to address. It is not automatically a missing keyword.

A deck may report 47 gaps. That is inventory, not priority. If your site never states a clear answer to one part of a buyer’s question, an engine may borrow that part from a competitor. If your existing pages are thin throughout, though, a long gap list will not give you a foundation.

Ask for **three buyer prompts you lose**, the missing passage in each, and the page that would change. That is a work plan. A topic count is not.

## Verified bot vs spoofed User-Agent

**A verified bot visit is a server hit checked against the bot’s identity, not just its claimed name.** Any scraper can put “Googlebot” or “OAI-SearchBot” in a User-Agent header. If a report counts headers alone, its bot traffic may include impostors.

[Google describes a reverse DNS check followed by a forward lookup confirming the same IP](https://developers.google.com/search/docs/crawling-indexing/verifying-googlebot); it also publishes IP ranges for verification. Apply that same verification discipline when reviewing other named bots.

Before paying for a bot-traffic audit, ask **how each visit was verified**. A convincing label in a log line is not proof of a visit.

## Live-answer fetch

**A live-answer fetch is a bot visit used to find material for a current answer, rather than to gather training data.** [OpenAI distinguishes OAI-SearchBot, used for search and citation discovery, from GPTBot, used for training collection](https://searchengineland.com/guide/ai-crawlers). Googlebot serves more than one search-related purpose, so the names and consequences differ by company.

The sales-deck version treats “AI bots” as one switch. That can produce bad instructions: block a training bot for privacy, then blame that block for a citation problem without checking the bot that handles search discovery. The mechanism matters before anyone touches the file.

Ask **which bot you are allowing or blocking, and which job that bot does**. Do not buy a citation fix based on the wrong User-Agent.

## Rendering and blocked resources

**Rendering is the process that makes page content available after its code loads.** What you see in a browser may not be what an answer engine’s crawler reads. [Many AI crawlers take the raw HTML without executing the JavaScript that injects content; Googlebot and Bingbot do render pages](https://searchengineland.com/guide/ai-crawlers).

If your pricing, reviews, or best answer appears only after a script runs, some citation bots may see a blank space where you see a polished section. That is a painful way to waste a good rewrite: the content is published, but the intended reader cannot reach it.

**View the page source**, then check whether the passage is present in the HTML. Do that before commissioning more content for the same page.

![Cutaway diagram of a website showing AI bots unable to read key content](https://groas.com/media/blog/e2375d84909d2a85aa6354c15d0bada2a77080e6be00238794afc387c3128f54.png)

## robots.txt and AI user agents

**`robots.txt` is a file that tells named crawlers which parts of a site they may access.** Its rules are not a master AI on-off switch. A line aimed at one User-Agent does not automatically govern every bot that can encounter your content.

[Blocking GPTBot addresses training collection, while live search discovery involves a different OpenAI bot](https://searchengineland.com/guide/ai-crawlers). Confuse those jobs and you can either leave access open where you meant to restrict it or restrict discovery where you wanted citations.

Ask the vendor to show the **actual `robots.txt` lines by bot name** and explain which ones affect training and which affect search discovery. “We handled AI crawling” is not an explanation.

## llms.txt

**`llms.txt` is a Markdown file at `/llms.txt` intended to summarize a site for models that choose to read it.** It is not a substitute for accessible pages or an established guarantee of citation. The sleight of hand is selling a tidy file as the thing that will fix an answer engine’s view of your business.

In one 90-day log of [62,100 AI bot visits, only 84 hits went to `/llms.txt`](https://otterly.ai/blog/the-llms-txt-experiment/), fewer than the average page in that log. A file is cheap to publish. Making it the centrepiece of an expensive AEO project is harder to defend.

Ask what share of **your verified citation-bot visits** fetched the file. Then put the serious work into readable passages and consistent business information.

## Prompt tracking

**Prompt tracking means repeatedly asking selected buyer questions and recording whether an engine names, links, or recommends you.** It is a sample, not a census of everything buyers might ask. Engine choice matters too: in one test of identical SaaS queries, GPT, Claude, and Gemini disagreed 54.5% of the time.

The useful question is not “Did the score rise?” It is **“What changed between runs?”** Ask which prompts were tracked, on which engines, how often they ran, and what changed on your site. Otherwise, the chart may tell you more about the sample than the work.

## Tracked prompt caps

**A tracked prompt cap is the number of questions a plan will monitor.** Plans may allow 25 prompts or 100, but a higher cap does not make the selection better. Fifty vanity questions can tell you less than ten questions buyers ask near a purchase decision.

I would rather watch the prompts where a buyer is comparing options than collect appearances in informational answers that never lead to a call. When a tracked question turns against you, the vendor should know what action follows. **Buy the response to a lost prompt, not the size of the prompt list.**

### Five questions to ask before paying for AEO

1. **Where do we lose?** Show me one buyer prompt, the passage a competitor supplies, and the page you would change. Not a gap count.
2. **What kind of appearance did we earn?** Split named, linked, and recommended results by surface: AI Overview, AI Mode, ChatGPT, or Perplexity.
3. **Can the bots read us?** Show the answer in page-source HTML, the relevant `robots.txt` lines, and verified fetch logs. [groas earned search](https://groas.com/earned-search) connects buyer-question and citation tracking to content rewrites and technical fixes instead of handing over a list of recommendations.
4. **Was the Overview appearance paid?** Label ads separately from citations. My ad spend is not evidence that the answer chose my page.
5. **What changed since the last score move?** If nothing on our site changed, check the prompts and engines before billing me to fix a falling chart.

## Ranking

**Ranking is a numbered position in classic search results, not a place inside an AI-generated paragraph.** Classic results still matter. Keep measuring them there. But an answer can name you, cite you, recommend you, or do none of those things, and a single position number cannot describe the difference.

I kept _ranking_ in AI reports longer than I should have because it made the meeting easier. It made the decision worse. For AI answers, I would retire the term and keep three counts instead: **how often buyers see your name, how often your URL supplies evidence, and how often the answer picks you**. Everything else is commentary.

## Frequently Asked Questions

### What is the difference between a mention, a citation and a recommendation in AI search?

A mention puts your brand name in the answer text without a link, a citation attaches your URL as evidence for a claim, and a recommendation tells the buyer you are the option to choose. Vendors often roll all three into one visibility score, which hides cases where your page supplies the evidence while a competitor gets the recommendation.

### Does share of voice in AI search mean my share of the whole market?

No. Share of voice in AI search is the percentage of tracked answers where your brand appears, and the tracked prompts are a sample the vendor chooses. Appearances in 15 of 40 prompts describe those 40 prompts, not a third of the market. Use it as a directional measure across the same prompts and engines, and ask to see the prompt list.

### Can a brand rank in an AI-generated answer the way it ranks in Google results?

No. An AI answer may name a brand in its verdict, link several sources, or leave other businesses out entirely, so there is no single numbered rank. An LLM response may cite only two to seven domains, and being one of those links still differs from being the company the sentence recommends. Ask whether you were named, linked, or recommended for each prompt.

### Are AI Overview, AI Mode and ChatGPT search the same thing?

No, they are different places for a buyer to encounter an answer. An AI Overview appears within Google's classic results page, where ordinary listings remain nearby. AI Mode is a separate conversational experience, and ChatGPT search and Perplexity sit outside Google with their own citation habits. Results should be split by surface before judging whether a visibility change matters.

### If my brand appears next to an AI Overview, does that mean the answer cited my site?

Not necessarily. An AI Overview ad is paid placement, not a citation: your brand can appear there because an ad was eligible to show, not because the answer chose your page as evidence. Ads can appear above, below, or within an AI Overview, and they remain ads. Ask the vendor to label every appearance paid or cited.

### What is answer-first content?

Answer-first content opens with a direct, self-contained answer before it expands, so a passage makes sense when lifted out of the page. A summary box pasted onto a vague article does not help, because a model looking for a quotable answer can use a competitor's clear passage instead. Ask which exact passage is meant to be cited and what was cut to make it stand alone.

### Why can't AI crawlers see content that appears when my page loads?

Many AI crawlers read the raw HTML without executing the JavaScript that injects content, so text that appears only after a script runs may be invisible to them, even though Googlebot and Bingbot do render pages. View the page source and check whether the passage is present in the HTML before commissioning more content for the same page.

### What is an entity in search, and why does it matter for my business?

An entity is a distinct business or thing with attributes and relationships a search system can recognize, such as a business's name, location, services, and connections to other sources. Entity work aims to reduce ambiguity in Google's Knowledge Graph. If business information conflicts across the site, Google Business Profile, and third-party listings, more blog posts will not fix that split identity.

## Related Posts

- [![Clay figure pours a wheelbarrow of articles into a wall funnel whose pipe is disconnected, so the pages pile on the floor beside an unused wrench.](https://groas.com/media/blog/e1a99963f159a4756e614472ee73400643c5900fa7bde718abe8fa563e8cdc77.png)](https://groas.com/post/can-chatgpt-even-read-your-site-a-45-min)
  ### [Can ChatGPT Read Your Site? A 45-Minute Check Before You Write More AEO Content](https://groas.com/post/can-chatgpt-even-read-your-site-a-45-min)
  [October 6, 2026 • 10 min read](https://groas.com/post/can-chatgpt-even-read-your-site-a-45-min)

  [Written by Alexander Perelman](https://groas.com/post/can-chatgpt-even-read-your-site-a-45-min)
- [![Cartoon of a marketer noting 'position two' after one pull of a slot machine whose reels are already spinning to a new result.](https://groas.com/media/blog/038bfefb4237637a1718ff66b379d5bdc42d57615f25797529aa5ef5682a4f7e.png)](https://groas.com/post/i-measured-ai-visibility-like-a-ppc-acco)
  ### [Six PPC Habits That Misread AI Visibility](https://groas.com/post/i-measured-ai-visibility-like-a-ppc-acco)
  [October 6, 2026 • 10 min read](https://groas.com/post/i-measured-ai-visibility-like-a-ppc-acco)

  [Written by David](https://groas.com/post/i-measured-ai-visibility-like-a-ppc-acco)
- [![A tiny worker on a ladder fixes up five model shopfronts while a giant dusty 'take a number' dispenser looms behind, its ticket queue ignored.](https://groas.com/media/blog/35c8ce6c06242ab08178711c5abe08dda0df51ec927a2dd6bc9770aa5f83bdbf.png)](https://groas.com/post/one-strategist-zero-developers-five-clie)
  ### [One Strategist, Zero Developers, Five Domains: A 30-Day AEO Launch Playbook](https://groas.com/post/one-strategist-zero-developers-five-clie)
  [October 6, 2026 • 10 min read](https://groas.com/post/one-strategist-zero-developers-five-clie)

  [Written by David](https://groas.com/post/one-strategist-zero-developers-five-clie)

## The Machines Already Run Search, You Should Probably Own One

[get started free](https://groas.typeform.com/to/xC1bQNUT)

## Structured data

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://graph.groas.com/entity/groas","name":"groas","url":"https://groas.com/","logo":{"@type":"ImageObject","url":"https://groas.com/icon-512.png","width":512,"height":512},"sameAs":["https://www.linkedin.com/company/groas/"]},{"@type":"WebSite","@id":"https://groas.com/#website","name":"groas","url":"https://groas.com/","inLanguage":"en-US","publisher":{"@id":"https://graph.groas.com/entity/groas"}},{"@type":"WebPage","@id":"https://groas.com/post/the-ai-search-glossary-for-owners-being#webpage","url":"https://groas.com/post/the-ai-search-glossary-for-owners-being","name":"The AI Search Glossary for Owners Being Pitched AEO","description":"A vendor says you have 38% share of voice in ChatGPT. Before you celebrate, ask whether that means a third of your market or appearances in 40 prompts…","inLanguage":"en-US","isPartOf":{"@id":"https://groas.com/#website"},"about":{"@id":"https://graph.groas.com/entity/groas"},"primaryImageOfPage":{"@type":"ImageObject","url":"https://groas.com/media/blog/49541cca7630a6a1b6238805ff04cb139b04352c33b943c873f05d03b6a04f72.png"},"breadcrumb":{"@id":"https://groas.com/post/the-ai-search-glossary-for-owners-being#breadcrumb"},"dateModified":"2026-10-06T05:23:57.870Z","mainEntity":{"@id":"https://groas.com/post/the-ai-search-glossary-for-owners-being#article"}},{"@type":"BreadcrumbList","@id":"https://groas.com/post/the-ai-search-glossary-for-owners-being#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://groas.com/"},{"@type":"ListItem","position":2,"name":"Blog","item":"https://groas.com/blog"},{"@type":"ListItem","position":3,"name":"AI For Google Ads","item":"https://groas.com/category/ai-for-google-ads"},{"@type":"ListItem","position":4,"name":"The AI Search Glossary for Owners Being Pitched AEO","item":"https://groas.com/post/the-ai-search-glossary-for-owners-being"}]},{"@type":"BlogPosting","@id":"https://groas.com/post/the-ai-search-glossary-for-owners-being#article","url":"https://groas.com/post/the-ai-search-glossary-for-owners-being","mainEntityOfPage":{"@id":"https://groas.com/post/the-ai-search-glossary-for-owners-being#webpage"},"isPartOf":{"@id":"https://groas.com/blog#blog"},"headline":"The AI Search Glossary for Owners Being Pitched AEO","description":"A vendor says you have 38% share of voice in ChatGPT. Before you celebrate, ask whether that means a third of your market or appearances in 40 prompts the vendor chose. Most AEO jargon is old PPC and SEO vocabulary with the meaning quiet…","datePublished":"2026-10-06T05:23:56.840Z","dateModified":"2026-10-06T05:23:57.870Z","image":{"@type":"ImageObject","url":"https://groas.com/media/blog/49541cca7630a6a1b6238805ff04cb139b04352c33b943c873f05d03b6a04f72.png"},"author":{"@id":"https://groas.com/author/david#person"},"publisher":{"@id":"https://graph.groas.com/entity/groas"},"articleSection":"AI For Google Ads","keywords":"Google Ads, AI For Google Ads","wordCount":2283,"inLanguage":"en-US"},{"@type":"Person","@id":"https://groas.com/author/david#person","name":"David","jobTitle":"Founder & CEO @ groas","image":"https://groas.com/media/blog/363be9ad5255816654dac065d01bf1c6063110b400bb38cc199570267232027c.jpg","worksFor":{"@id":"https://graph.groas.com/entity/groas"}},{"@type":"FAQPage","@id":"https://groas.com/post/the-ai-search-glossary-for-owners-being#faq","mainEntity":[{"@type":"Question","name":"What is the difference between a mention, a citation and a recommendation in AI search?","acceptedAnswer":{"@type":"Answer","text":"A mention puts your brand name in the answer text without a link, a citation attaches your URL as evidence for a claim, and a recommendation tells the buyer you are the option to choose. Vendors often roll all three into one visibility score, which hides cases where your page supplies the evidence while a competitor gets the recommendation."}},{"@type":"Question","name":"Does share of voice in AI search mean my share of the whole market?","acceptedAnswer":{"@type":"Answer","text":"No. Share of voice in AI search is the percentage of tracked answers where your brand appears, and the tracked prompts are a sample the vendor chooses. Appearances in 15 of 40 prompts describe those 40 prompts, not a third of the market. Use it as a directional measure across the same prompts and engines, and ask to see the prompt list."}},{"@type":"Question","name":"Can a brand rank in an AI-generated answer the way it ranks in Google results?","acceptedAnswer":{"@type":"Answer","text":"No. An AI answer may name a brand in its verdict, link several sources, or leave other businesses out entirely, so there is no single numbered rank. An LLM response may cite only two to seven domains, and being one of those links still differs from being the company the sentence recommends. Ask whether you were named, linked, or recommended for each prompt."}},{"@type":"Question","name":"Are AI Overview, AI Mode and ChatGPT search the same thing?","acceptedAnswer":{"@type":"Answer","text":"No, they are different places for a buyer to encounter an answer. An AI Overview appears within Google's classic results page, where ordinary listings remain nearby. AI Mode is a separate conversational experience, and ChatGPT search and Perplexity sit outside Google with their own citation habits. Results should be split by surface before judging whether a visibility change matters."}},{"@type":"Question","name":"If my brand appears next to an AI Overview, does that mean the answer cited my site?","acceptedAnswer":{"@type":"Answer","text":"Not necessarily. An AI Overview ad is paid placement, not a citation: your brand can appear there because an ad was eligible to show, not because the answer chose your page as evidence. Ads can appear above, below, or within an AI Overview, and they remain ads. Ask the vendor to label every appearance paid or cited."}},{"@type":"Question","name":"What is answer-first content?","acceptedAnswer":{"@type":"Answer","text":"Answer-first content opens with a direct, self-contained answer before it expands, so a passage makes sense when lifted out of the page. A summary box pasted onto a vague article does not help, because a model looking for a quotable answer can use a competitor's clear passage instead. Ask which exact passage is meant to be cited and what was cut to make it stand alone."}},{"@type":"Question","name":"Why can't AI crawlers see content that appears when my page loads?","acceptedAnswer":{"@type":"Answer","text":"Many AI crawlers read the raw HTML without executing the JavaScript that injects content, so text that appears only after a script runs may be invisible to them, even though Googlebot and Bingbot do render pages. View the page source and check whether the passage is present in the HTML before commissioning more content for the same page."}},{"@type":"Question","name":"What is an entity in search, and why does it matter for my business?","acceptedAnswer":{"@type":"Answer","text":"An entity is a distinct business or thing with attributes and relationships a search system can recognize, such as a business's name, location, services, and connections to other sources. Entity work aims to reduce ambiguity in Google's Knowledge Graph. If business information conflicts across the site, Google Business Profile, and third-party listings, more blog posts will not fix that split identity."}}]}]}
```