# What AI Engines Cite: A Source Map From 27,007 Citations

> We classified 27,007 citations from 4,938 AI answers across ten engines. The brand's own site earned 0.9% of them; decision pages earned 40.1%.

Source: https://mentionx.ai/learn/research/what-ai-engines-cite  
Author: Dipuck Jones, founder of MentionX  
Updated: 12 November 2026  
Level: Practitioner

**We analysed 27,007 citations attached to 4,938 AI answers to category questions across ten AI engines.** The brand's own website earned 0.9% of them. Pages built to help buyers decide earned 40.1%, and each engine draws on a visibly different part of the web.

## Abstract

When a buyer asks an AI engine which product to choose, the answer is assembled from sources the engine retrieves and, often, cites. We analysed 27,007 citations attached to 4,938 AI answers to category questions (questions that do not name the brand being studied) across ten AI engines and answer surfaces. We classified every cited domain by source type and every cited URL by page type, and compared the patterns across engines and across two groups of software categories.

Three results stand out. First, the brand's own website accounted for 0.9% of citations on category questions. Second, engines draw on visibly different parts of the web: Google AI Mode took 17.5% of its citations from community and video platforms, while ChatGPT did not cite Reddit once in any of its 14,590 citations in the dataset, whether or not the question named a brand. Third, pages built to help buyers decide (best-of lists, alternatives, comparisons and reviews) made up 40.1% of all visible citations.

## Key findings

- **0.9%** of citations on category questions pointed to the brand's own website. 99.1% pointed somewhere else.
- **40.1%** of visible citations went to decision pages: best-of lists (24.9%), alternatives pages (7.5%), comparisons (4.4%) and reviews (3.3%).
- **17.5%** of Google AI Mode's citations came from community and video platforms such as Reddit and YouTube. In ChatGPT's answers, that share was 0.0%.
- **35.1%** of Gemini's citations were Google grounding links that hide the original source, making them impossible to attribute from the answer alone.
- **56.3%** of answers carried at least one citation, with an average of 9.7 citations per cited answer.
- **15.9%** of citations were linked by the engine to a specific brand in the answer. Most citations support the answer as a whole, not a named product.

## Why this matters

When a buyer asks an AI engine which product to choose, the engine answers with a shortlist and, often, the pages it read to build it. Those pages decide which names reach the shortlist. If a brand's own site supplied most of them, marketing teams would already control the input. This study shows that it does not: 99.1% of citations on category questions point somewhere else.

Knowing which pages, on which sites, each engine reads is therefore the starting point for any plan to appear in AI answers. This report maps those sources across ten engines.

## 1. Research question

AI engines increasingly act as the first shortlist in a software purchase. For marketers, the practical question is not only whether a brand is mentioned, but which sources the engine relies on when it decides what to say. If engines cite a brand's own pages, owned content is the lever. If they cite third parties, the lever is coverage elsewhere.

This study asks four questions:

1. How often do AI engines cite sources on category questions, and how many?
2. What kinds of sources do they cite?
3. Do engines differ in where they look?
4. What kinds of pages get cited?

## 2. Data and method

**Answers.** The dataset contains answers from MentionX scans of B2B software brands. Each scan sends a fixed set of buyer-style prompts to each selected engine and stores every answer with its citations.

**Category questions only.** A prompt is excluded when its text contains the scanned brand's name as a whole word. Questions that name a brand are answered differently: engines cite the brand's own pages heavily, which would distort a study of how buyers discover options. After exclusion, 4,938 answers and 27,007 citations remain.

**Engines.** ChatGPT, Gemini, Google AI Mode, Google AI Overviews, Perplexity, Microsoft Copilot, Claude, Grok, DeepSeek and Meta AI.

**Category groups.** Results are reported for all answers and for two groups: SaaS categories (such as CRM, workflow software and real estate technology) and AI visibility tools.

**Source types.** Each cited domain is normalised (lower case, no "www.") and assigned to a source type using fixed domain lists: tech media, review sites, analysts, community, video, social and publishing, reference, press wires, the scanned brand's own site, and hidden Google links. Domains not on any list are counted as "other". The full lists are in the appendix.

**Page types.** Each visible citation URL is assigned a page type from patterns in its address: best-of list, alternatives, comparison, review, pricing, documentation or guide, blog or article, homepage, or other page.

**Visible citations.** All citations except hidden Google links, whose real source cannot be read from the answer. Shares by engine, by page type and by domain are calculated on visible citations; shares by source type are calculated on all citations.

How a MentionX scan collects and reads answers, and how every KPI is defined, is set out in [How MentionX turns AI answers into evidence](https://mentionx.ai/learn/research/ai-answer-methodology-and-dataset). The general approach to ranges and comparability is on the [methodology page](https://mentionx.ai/learn/methodology).

## 3. How often engines cite

Of 4,938 answers, 56.3% carried at least one citation. Answers that cited anything cited a lot: 9.7 sources on average.

| Engine | Answers with a citation | Citations per cited answer |
|---|---|---|
| Grok | 100.0% | 8.0 |
| Google AI Mode | 83.4% | 5.0 |
| ChatGPT | 75.0% | 10.3 |
| Google AI Overviews | 69.0% | 6.2 |
| Perplexity | 66.7% | 10.4 |
| Microsoft Copilot | 47.2% | 4.3 |
| Gemini | 42.5% | 11.5 |
| Meta AI | 31.6% | 46.9 |
| Claude | 23.9% | 11.4 |
| DeepSeek | 0.3% | n/a |

Engines differ in style as much as in frequency. Google AI Mode and Copilot cite in most answers but list few sources. Meta AI cites less often but, when it does, lists dozens of sources. DeepSeek returned almost no citations in the answers we collected, so it is left out of the source analysis below. Grok, Meta AI and Claude contribute fewer answers to the dataset than the larger engines, so their rates are less precise.

## 4. Where citations come from

| Source type | Share of all citations |
|---|---|
| Other domains (vendor sites, niche publishers, blogs) | 81.7% |
| Hidden Google links | 8.8% |
| Tech media | 2.3% |
| Social and publishing | 1.3% |
| Review sites | 1.3% |
| Video | 1.2% |
| Community | 1.2% |
| Brand's own site | 0.9% |
| Analysts | 0.5% |
| Reference (Wikipedia) | 0.4% |
| Press wires | 0.3% |

Three observations follow.

**The brand's own site is a small voice.** On category questions, 0.9% of citations pointed to the scanned brand's domain, about one in every 108. In the SaaS group, where brands are more established, the share was 1.7%. In the AI visibility tools group, it was 0.2%.

**Hidden sources are not trivial.** 8.8% of all citations were Google grounding links (vertexaisearch.cloud.google.com) that redirect to the real source without naming it. All of them came from Gemini, where they made up 35.1% of citations. Any analysis of Gemini's sources that reads only the answer will miss a third of them.

**The long tail dominates.** The well-known source classes, from tech media to press wires, together account for fewer than one in ten citations. More than four in five go to domains no standard source list covers: vendor sites, specialist publishers and category blogs. This is a finding, not a gap in the method. The material AI engines draw on is far wider than the review sites and analysts most marketers track.

## 5. Engines look in different places

The source mix varies sharply by engine. Shares below are of each engine's visible citations.

| Engine | Community and video | Tech media, reviews and analysts | Brand's own site |
|---|---|---|---|
| Google AI Mode | 17.5% | 2.9% | 1.3% |
| Google AI Overviews | 15.4% | 0.1% | 0.0% |
| Meta AI | 3.7% | 2.6% | 0.0% |
| Gemini | 3.2% | 0.6% | 1.2% |
| Perplexity | 0.5% | 0.8% | 0.1% |
| ChatGPT | 0.0% | 8.2% | 1.5% |
| Microsoft Copilot | 0.0% | 2.8% | 0.0% |

Google's answer surfaces lean on Reddit and YouTube. ChatGPT does not: across every answer we collected, including those to questions that name a brand, ChatGPT cited Reddit zero times in 14,590 citations, and YouTube accounted for under 0.1%. Instead, ChatGPT gave tech media, review sites and analysts a larger share than any other engine.

The practical consequence is that there is no single source strategy for "AI search". A brand that is discussed on Reddit and YouTube but absent from tech media and review sites may be visible in Google's AI answers and invisible in ChatGPT's, and the reverse.

## 6. What kind of page gets cited

| Page type | Share of visible citations |
|---|---|
| Other pages | 32.8% |
| Best-of lists | 24.9% |
| Blog posts and articles | 13.8% |
| Homepages | 8.4% |
| Alternatives pages | 7.5% |
| Comparisons | 4.4% |
| Reviews | 3.3% |
| Documentation and guides | 2.9% |
| Pricing pages | 2.0% |

Decision pages (best-of lists, alternatives, comparisons and reviews) made up 40.1% of visible citations. The share was higher in the AI visibility tools group (47.5%) than in the SaaS group (31.2%), driven by best-of lists (29.1% against 19.9%) and alternatives pages (10.6% against 3.7%).

This matches how the questions are asked. A buyer asking "which tools are best for…" gets an answer built from pages that already rank and compare tools. Being included on those pages, accurately, matters more than any single page a brand publishes about itself.

## 7. A long tail, with a few dominant voices

Citations are spread across thousands of domains, and most domains appear rarely.

| | SaaS categories | AI visibility tools |
|---|---|---|
| Share of visible citations held by the top 10 domains | 11.6% | 16.8% |
| Share held by the top 100 domains | 36.1% | 49.5% |
| Share of domains cited only once | 58.5% | 48.1% |

In the SaaS group, the most-cited domains were TechRadar (3.5% of visible citations), google.com (1.9%), Gartner (0.9%), Semrush (0.9%), Wikipedia (0.8%), YouTube (0.8%), G2 (0.7%), Reddit (0.7%), SoftwareConnect (0.6%) and HubSpot (0.6%). Domains belonging to the scanned brands are excluded from this list.

The AI visibility tools group is more concentrated, and its leaders are different in kind: seven of its ten most-cited domains belong to vendors in that category. In a young category with little independent coverage, vendors' own comparison and best-of content becomes the material engines cite.

## 8. Citations support answers more than brands

Engines link some citations to a specific brand named in the answer. Only 15.9% of citations were linked in this way to a tracked brand, meaning the studied brand or one of the competitors its scan follows. The rest support the answer in general: market overviews, buying guides and lists. For a brand, this means visibility in AI answers depends heavily on being present in the general material engines retrieve about a category, not only in pages about the brand itself.

## 9. Example: one question, five engines

To show what these figures look like in a single answer, we took one category question from the dataset, asked about product walkthrough software: *"What tools help sales teams share a self-serve product walkthrough after a demo call?"* Five engines answered it with citations. Vendor names are withheld.

| Engine | Citations | What the engine cited |
|---|---|---|
| Gemini | 14 | Six best-of lists and five other vendor blog posts about demo software, plus a Reddit thread, a YouTube video and an analyst review page |
| Perplexity | 10 | All ten from blogs of vendors in the category, four from a single vendor; seven were best-of lists |
| ChatGPT | 8 | Two best-of lists, vendor and CRM blogs, a homepage, a Japanese blog post and one page unrelated to the question |
| Microsoft Copilot | 5 | Vendor blog posts, an alternatives page, a product directory and a software catalogue |
| Google AI Mode | 2 | One best-of list and one YouTube video |

Of 39 citations, only one article was cited by more than one engine. The brand whose scan asked the question was not cited by any of them. Most cited pages were written by competitors in the same category, and most were titled as lists of the best tools. This is the pattern of sections 4 to 7 in miniature: each engine built its answer from a different set of pages, mostly decision pages, mostly from the long tail.

## 10. Our view: why this happens

The figures above are measured. This section is our interpretation.

**Engines borrow other people's judgement.** On a "which tool" question, an engine prefers pages that have already compared the options. A product page describes one option; a best-of list describes the market. That is why decision pages earn 40.1% of visible citations and the brand's own site 0.9%.

**In many categories, competitors write the lists.** Where independent coverage is thin, vendors' comparison and best-of posts fill the gap: seven of the ten most-cited domains in the AI visibility tools group belong to vendors, and in the example above most cited pages were written by rivals. The vendor that publishes the most useful list gets a voice in its competitors' answers.

**Each engine has its own reading habit.** Google's answer surfaces read the community web; ChatGPT reads editorial and review sites. These differences most likely come from different search indexes and retrieval choices, so they are unlikely to disappear on their own.

**The long tail is an opening.** Because 81.7% of citations go to domains outside the usual lists, a specialist publisher or a well-made category guide can be cited as often as a major review site.

## 11. What marketers can do

1. **Map sources per engine, for your own questions.** Run the category questions your buyers ask on each engine, collect the citations, and sort them by source type and page type. A single list of "AI sources" hides the differences between engines.
2. **Get onto the decision pages already being cited.** List the best-of, alternatives, comparison and review pages cited for your category. Check whether you appear on each one, and whether your description, pricing and use cases are correct. Ask authors to add or correct entries, with evidence.
3. **Publish the comparison content your category lacks.** Where rivals' lists are the main source, publish fair, specific comparison and alternatives pages with clear criteria and dated facts. In categories with little independent coverage, these pages become source material.
4. **Match each engine's reading habit.** For Google AI Mode and AI Overviews, take part in the Reddit and YouTube conversations about your category. For ChatGPT, invest in coverage from tech media, review sites and analysts.
5. **Keep owned pages as the reference.** Owned pages matter when buyers ask about the brand by name. Make pricing, features and integrations easy to read, so an engine has a correct source when it gets there.
6. **Account for hidden sources.** Resolve Gemini's grounding links before drawing conclusions about its sources. Otherwise, about a third of them stay invisible.
7. **Re-check over time, and plan for the long tail.** Answers and citations change between runs, and most citations go to domains outside any well-known list. Track the source map for your category over time, not once.

## 12. Limitations

- **Sample.** The dataset covers the categories MentionX has scanned. It is not a census of the web or of all software categories.
- **Category groups are unequal.** The AI visibility tools group is smaller and more concentrated than the SaaS group; its figures should be read with that in mind.
- **Answers vary.** AI answers can differ by user, location and time. Results describe these answers, not every person's experience.
- **Generated prompts.** Prompts are generated from each brand's website and market, not sampled from real buyer searches.
- **Rule-based classification.** Source types use fixed domain lists, and page types use URL patterns. Both are transparent but imperfect: a best-of article with a plain URL is counted as "other page".
- **Citation is not influence.** A citation shows what an engine linked, not how much a source shaped the answer.
- **Engine coverage.** DeepSeek returned almost no citations, and some engines contribute fewer answers than others.

## How to cite this report

Jones, D. (2026). *What AI engines cite: a source map from 27,007 citations.* MentionX Research. mentionx.ai/learn/research/what-ai-engines-cite

## Data requests

Journalists and researchers can request the aggregated tables behind this report at press@mentionx.ai. Figures may be quoted with attribution to MentionX Research.

## Appendix: classification rules

**Source types (by domain).** Community: reddit.com, quora.com, news.ycombinator.com. Video: youtube.com. Social and publishing: linkedin.com, x.com, twitter.com, facebook.com, medium.com. Review sites: g2.com, capterra.com, softwareadvice.com, getapp.com, trustradius.com, sourceforge.net, softwarereviews.com, softwareconnect.com, peerspot.com. Analysts: gartner.com, forrester.com, idc.com. Reference: wikipedia.org. Press wires: prlog.org, openpr.com, prnewswire.com, businesswire.com, globenewswire.com. Tech media: techradar.com, pcmag.com, zdnet.com, forbes.com, techcrunch.com, cnet.com, tomsguide.com, businessinsider.com. Hidden Google links: vertexaisearch.cloud.google.com. Brand's own site: the scanned brand's domain and its subdomains.

**Page types (by URL, first match wins).** Comparison: "vs", "versus" or "compar". Alternatives: "alternative". Best-of list: "best" or "top" as a path word. Review: "review". Pricing: "pricing" or "/price". Documentation or guide: docs, help, support, knowledge, learn or guide paths. Blog or article: blog, insights, resources, article or news paths. Homepage: a domain with no path.

## Frequently asked questions

### Does a brand's own website matter for AI answers?

Yes, but mainly when buyers ask about the brand by name. On category questions, the brand's own site made up 0.9% of citations.

### Which sources do AI engines cite most?

Mostly vendor sites, specialist publishers and blogs. Among named source types, tech media, social platforms, review sites, video and community sites each held between 1% and 2.5% of citations.

### Do ChatGPT and Google cite the same sources?

No. Google AI Mode took 17.5% of its citations from community and video platforms; ChatGPT cited Reddit zero times across all 14,590 of its citations in the dataset, including answers to questions that name a brand, and gave tech media, review sites and analysts a larger share.

### What type of page is cited most?

Best-of lists, at 24.9% of visible citations. Together with alternatives pages, comparisons and reviews, decision pages made up 40.1%.

---

Written and reviewed by Dipuck Jones for MentionX Learn. The HTML version of this page: https://mentionx.ai/learn/research/what-ai-engines-cite