The discovery layer has collapsed into five answer engines. Five operators. Five proprietary model families. Five distinct citation logics. The buyer asks the question. The engine answers it. Whichever brand is named in the answer enters the consideration set. Whichever brand is not is invisible.

Citation Share is the new market share. Not impressions. Not clicks. Not rankings. Share of the answer.

More than a third of consumers now begin product research with AI, not Google. The number is rising every quarter. For B2B buyers, the share is higher. For software, professional services, and category research, AI is already the dominant entry point — and the buyer never sees the brands the engine did not name.

What Is an Answer Engine

A search engine returns a list of links. The user picks the source. A chatbot generates conversational text from a fixed training set — no retrieval, no citation. An answer engine retrieves from a live source set, synthesizes a single answer, and cites a small number of authoritative sources. The user gets one answer. The brands cited are the brands the buyer considers. The rest is invisible. Five engines mediate the category at scale.

The Five Engines

1. ChatGPT — the default engine.
2. Claude — the editorial engine.
3. Gemini — the workspace engine.
4. Perplexity — the citation engine.
5. Google AI Overviews — the search-surface engine.

The Side-by-Side Index

Field	ChatGPT	Claude	Gemini	Perplexity	AI Overviews
Operator	OpenAI	Anthropic	Google DeepMind	Perplexity AI	Google
Founded	2015 (Nov 2022 launch)	2021	2023	2022	2024
Users	800M+ weekly	Hundreds of millions monthly	750M+ monthly	~100M MAU + Sonar API	Inside Google Search
Model family	GPT-5, o-series, GPT-4o	Claude Opus 4.x, Sonnet 4.6, Haiku 4.5	Gemini 3, 2.5 Pro/Flash	Routes across GPT-5, Claude, Sonar, Grok	Gemini-powered
Top sources	Wikipedia, Reddit, AP, WSJ, FT, Vox, Condé Nast, Time	NYT, Atlantic, New Yorker, Economist, Guardian, WIRED, Reuters	YouTube, Reuters, Forbes, FT, Google properties, Wikipedia	Reuters, AP, Bloomberg, WSJ, trades	Reddit (~21%), YouTube (~19%), Wikipedia, Google properties
Reddit weight	High	Lowest of any major engine	Moderate	Moderate	Highest of any engine (~21%)
YouTube weight	Low	Low	High	Moderate	Highest of any engine (~19%)
Inline citations	In Search mode	Selective	In Overviews and AI Mode	Always — defining feature	Always
Linked playbook	ChatGPT	Claude	Gemini	Perplexity	AI Overviews

Engine Profiles

ChatGPT — The Default Engine

OpenAI's product. Launched November 30, 2022. Largest by direct reach with over 800 million weekly active users. The default name in the category. Citation profile: Wikipedia dominates entity queries at 26–48% of top-source share. Reddit weighted heavily on opinion and recommendation. Publisher deal stack — AP, News Corp, FT, Axel Springer, Vox, Condé Nast, Hearst, Time, Le Monde, Prisa, Dotdash Meredith — materially over-indexed. NYT lawsuit ongoing; NYT content still surfaces but under-weighted. The ranked authority insight: ChatGPT rewards brands that show up in the publisher deal stack and in Wikipedia simultaneously. Full playbook: How to Rank on ChatGPT in 2026.

Claude — The Editorial Engine

Anthropic's product. Founded 2021 by Dario and Daniela Amodei. Most distinctive citation profile of any major engine — over-indexed on The New York Times, The Atlantic, The New Yorker, The Economist, The Guardian, The Washington Post, WIRED, and Reuters. Reddit weighted lower than on any other engine. Enterprise stack chose it — Apple Intelligence routes to it, Amazon Bedrock features it, Salesforce Einstein uses it, Pfizer, Notion, Zoom deploy it. Constitutional AI training. No-ads commitment. ASL-3/ASL-4 safety framework. Project Glasswing runs the Mythos Preview frontier model. The ranked authority insight: only 36% of Claude's journalism citations are under twelve months old versus 56% on ChatGPT — content that lasts compounds harder here than on any other engine. Full playbook: How to Rank on Claude in 2026.

Gemini — The Workspace Engine

Google DeepMind's product. 750 million monthly active users. Embedded across Gmail, Docs, Sheets, Slides, Workspace, Android, Chrome, NotebookLM. Reach by distribution, not by user choice. YouTube heavily over-indexed. Google Maps, Scholar, Business Profile for local, academic, and brand queries. Knowledge Graph is the entity layer, fed by Wikipedia, Wikidata, LinkedIn, Crunchbase, Google Business Profile. The ranked authority insight: top-10 organic Google ranking is the strongest single input. Brands without organic search authority cannot win Gemini citation. Full playbook: How to Rank on Gemini in 2026.

Perplexity — The Citation Engine

Founded 2022. Approximately $18 billion valuation. Backers include institutional venture, NVIDIA, Jeff Bezos personally, and sovereign-aligned funds. Built around inline citation as the defining feature. Commercial bet is Sonar — the search-grounded LLM API for developers and enterprises — plus the Bharti Airtel India deal giving 360 million subscribers free Pro access. Active litigation from Dow Jones, NYT, Forbes, and a Japanese publisher consortium. The ranked authority insight: Perplexity multi-models internally. Win on Claude, win on GPT-5, and Perplexity surfaces the brand because it is routing through those models. Full playbook: How to Rank on Perplexity.

Google AI Overviews — The Search-Surface Engine

Powered by Gemini. Lives inside Google Search results pages. Appears on a substantial and growing share of U.S. Google queries. Largest AI-mediated impression surface by reach. Reddit at ~21% top-source share (anchored by Google's reported $60M annual data deal with Reddit), YouTube at ~19%, plus Wikipedia, Google-owned properties, and major editorial. Top-10 organic Google ranking is the floor. The regulatory layer matters: DOJ Search remedies ongoing, EU Digital Markets Act compliance pressure, publisher backlash with double-digit traffic declines reported. The ranked authority insight: zero-click is the operating reality. Optimize for citation share inside the Overview, not just click-through. Full playbook: How to Rank on Google AI Overviews and AI Mode in 2026.

What Is True Everywhere

• Wikipedia is foundational. Every engine weights it heavily on entity queries.
• Schema markup is the parsing layer. Article, Organization, Person, FAQPage, Product, VideoObject.
• Named authors outperform anonymous content. Person schema with verifiable credentials.
• Original research with named data is the most-cited content type. Studies, indexes, benchmarks.

What Is Engine-Specific

• Reddit weight — heavy on ChatGPT and AI Overviews. Lowest on Claude.
• Recency preference — ChatGPT skews recent. Claude rewards evergreen.
• Video weight — Google-stack lever (Gemini, AI Overviews), not universal.
• Publisher preference — ChatGPT favors the deal stack, Claude favors editorial depth, Gemini and AI Overviews favor Reuters, Forbes, FT, Time.

The Citation Share Approach

The same measurement discipline works across all five engines. The methodology is the Citation Share Index, with source-level analysis in the AI Platform Citation Source Index 2026.

1. Build a frozen prompt set — 15 to 50 prompts mirroring real buyer queries.
2. Run the same prompts every two weeks across all five engines.
3. Log three metrics per prompt per engine: citation frequency, share of model, source mix.
4. Report monthly. Reallocate quarterly.

What This Means for PR Firms

The buyer expects coverage on all five engines. The earned-only firm cannot deliver it. The new requirement is integrated: earned media, plus Generative Engine Optimization, plus AI Visibility research, plus Knowledge Graph optimization, plus video, plus community presence, plus original data publishing. One operating system. One firm. The argument is laid out in full at Public Relations Changes. 5W is the AI Communications Firm — building brand authority across ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews. Full positioning at AI Communications.

The Challengers — Grok and Meta AI

Grok — xAI's product, embedded in X. Measurable citation share on news and current-event queries. Under-indexes mainstream editorial, over-indexes X posts and community signal. Not yet in the five — consumer reach and enterprise integration remain materially smaller.

Meta AI — embedded inside Instagram, WhatsApp, and Facebook. Over 600 million monthly active users by total platform reach. Citation profile still emerging. Meta has not pushed it as a buyer-research surface. Not yet in the five because it does not function as a primary discovery engine for commercial intent. Both could be in two years.

The Cluster

• How to Rank on ChatGPT in 2026
• How to Rank on Claude in 2026
• How to Rank on Gemini in 2026
• How to Rank on Perplexity
• How to Rank on Google AI Overviews and AI Mode in 2026
• Comparison Hub
• Citation Share Index
• AI Platform Citation Source Index 2026
• GEO
• AI Communications

---

Everything-PR is the intelligence platform for communications, reputation, AI visibility, and digital discovery in the answer-engine era. Thirty-plus publications. Publishing since 2009. Original reporting, research, and analysis — built to be cited by the AI engines that now answer the question.

Frequently Asked Questions

What is an answer engine?

A system that retrieves from a live source set, synthesizes one answer, and cites a small number of authoritative sources. Different from a search engine (which returns a link list) and a chatbot (which generates from fixed training data).

What are the five answer engines?

ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews.

Why is ChatGPT first?

Largest direct reach with 800M+ weekly active users. The default name in the category.

Why is Claude called the editorial engine?

Most distinctive citation profile — over-indexed on long-form editorial (NYT, Atlantic, New Yorker, Economist), under-indexed on Reddit, with the highest evergreen content weighting of any major engine.

Why is Gemini called the workspace engine?

Embedded across Gmail, Docs, Sheets, Workspace, Android, and Chrome. Reach by distribution, not by direct user choice.

Why is Perplexity called the citation engine?

Inline citation is the defining product feature — every answer shows sources at the point of consumption.

Why is Google AI Overviews listed separately from Gemini?

Both run on the Gemini model family, but they are different surfaces with different ranking inputs. AI Overviews lives inside Google Search and is downstream of top-10 organic ranking. Gemini direct is a standalone assistant.

What is Citation Share?

The share of answers across the five engines in which the brand is named. The new market share for the AI-mediated buyer journey.

How do I measure Citation Share?

Frozen prompt set, run every two weeks across all five engines, logging citation frequency, share of model, and source mix.

What patterns hold across all five engines?

Wikipedia is foundational. Schema markup is essential. Named authors outperform anonymous bylines. Original research is the most-cited content type.

Where do the engines diverge?

Reddit weight, recency preference, video weight, and publisher preference. Strategy must diverge with them.

What about Grok and Meta AI?

Both have measurable share but are not in the operating set today. Grok matters most for news and current-event categories. Meta AI matters most for B2C reach scale.

How much does AI-mediated discovery already account for?

More than a third of consumers now begin product research with AI, not Google. Higher for B2B. The number rises every quarter.

Who handles AI Communications across all five engines?

5W is the AI Communications Firm. Citation Share is the new market share.