Commercial AI search behaves differently from informational search. When a user asks an AI system to compare options, recommend providers, or identify the best solution for a need, the sources it cites shift toward content that helps evaluate choices rather than content that simply explains a topic.
A large Wix Studio AI Search Lab and Peec AI study makes that difference unusually visible. Across more than one million citations, commercial-intent prompts disproportionately surfaced listicles, category pages, and discussion content—while traditional articles and product pages played much smaller roles.
The practical lesson is not “publish more listicles.” It is more useful than that: match the content format to the decision the buyer is asking AI to help make.
Commercial AI queries favor evaluation content over explanation content.
When the job shifts from “help me understand” to “help me choose,” the citation mix shifts with it.
What the 1,056,727-Citation Study Actually Analyzed
The Wix Studio AI Search Lab / Peec AI research on content types cited by LLMs analyzed 75,000 AI answers containing 1,056,727 citations. The data came from ChatGPT, Google AI Mode, and Perplexity and was segmented across four prompt intents and five industries.
A large citation dataset rather than a handful of anecdotal prompts.
Answers were classified by user intent to expose format differences.
ChatGPT, Google AI Mode, and Perplexity were included.
eCommerce, health and wellness, home repair, professional services, and SaaS.
The study classified prompts as informational, commercial, navigational/local, or transactional. Commercial prompts represented users researching products or services before making a final choice.
What Counts as Commercial Content in AI Search?
Commercial intent sits between learning and acting. The user has moved beyond a broad educational question, but has not necessarily chosen a vendor or product yet.
Prompts such as “best payroll software for a 50-person company,” “top accounting firms for SaaS startups,” or “which project management tools are best for agencies?” ask the AI system to organize a market, compare alternatives, and reduce a decision set.
That matters because the best source for explaining a category is not always the best source for choosing within it.
“Which options should I seriously consider?”
Commercial content earns its value by helping answer that question with categories, alternatives, tradeoffs, evidence, and comparative context.
The Citation Mix Changes Dramatically for Commercial Queries
For commercial-intent prompts, listicles accounted for 40.86% of citations. Category pages were second at 12.42%, followed closely by discussions at 11.44%.
Traditional articles, by contrast, represented only 6.15% of commercial citations. Product pages accounted for 7.14%.
| Content type | Share of commercial citations | What the format helps a buyer do |
|---|---|---|
| Listicle | 40.86% | Scan, compare, shortlist, and understand category options. |
| Category page | 12.42% | Browse a structured set of relevant products or services. |
| Discussion | 11.44% | Access peer opinions, lived experience, and market sentiment. |
| Product page | 7.14% | Evaluate a specific offering once interest is established. |
| Other | 6.37% | Capture specialized source types outside the main classifications. |
| Article | 6.15% | Explain a topic or provide broader educational context. |
| Profile | 5.41% | Provide basic entity or provider information. |
| Comparison | 4.08% | Evaluate a narrower set of named alternatives directly. |
| How-to guide | 3.89% | Explain process rather than choose a provider. |
| Homepage | 1.70% | Confirm company identity, not usually solve comparison intent. |
| Alternative page | 0.54% | Address a specific replacement or competitor scenario. |
What the Data Tells Us About Commercial AI Search
Evaluation beats explanation
Articles dominate informational intent, but they fall to 6.15% of commercial citations. Once a buyer asks AI to choose, structured decision content becomes much more useful.
Commercial does not mean transactional
Product pages are strong when the buyer is ready to act, but only 7.14% of commercial citations. Research-stage buyers still need the market organized before they need the checkout path.
Third-party context can matter
Commercial questions often require comparisons that brand-owned product pages cannot independently provide. This helps explain why recommendation-oriented and peer-driven formats can surface heavily.
Intent matters more than the acronym
The study found prompt intent was more predictive of cited content type than either industry or AI model. The buyer task should drive the content strategy first.
Commercial Search Is a Different Stage of the Buyer Journey
One of the easiest mistakes in GEO is to treat every prompt as a variation of traditional informational search. The study suggests a more useful model.
When someone asks an informational question, AI systems can lean heavily on educational articles. When the same person moves into commercial research, the system needs sources that organize choices. When the person becomes transactional, product and category pages become more useful again.
That creates a simple content journey:
| Buyer stage | Typical AI job | Content most useful to the decision |
|---|---|---|
| Informational | Explain the topic | Articles, how-to guides, educational resources |
| Commercial | Organize and compare options | Best Of lists, category pages, discussions, reviews, comparisons |
| Transactional | Help the user act | Product pages, service pages, category pages, conversion paths |
A modern search strategy therefore needs more than “great blog content.” It needs content for each decision stage—and a plan for visibility in the third-party sources that buyers and AI systems may use to compare the market.
Industry Matters, but It Does Not Override Intent
The research did find meaningful industry differences across the full dataset. SaaS showed the strongest overall bias toward listicles at 35.37%. Professional services followed at 25.24%. Health and wellness leaned more heavily toward articles, while eCommerce showed a more balanced mix of listicles, articles, and category pages.
Those differences are useful, but they do not overturn the broader finding: what the user is trying to accomplish is the first variable to solve.
A SaaS company should not conclude that every page needs to be a listicle. A healthcare brand should not conclude that educational articles are enough for every commercial question. The more practical approach is to map the prompt intent first, then refine the content mix for the industry.
What Marketers Should Do With This Research
The study gives marketers a better planning system than “publish more content.” Start with the buyer's AI question, then decide what evidence and format the answer requires.
Map your commercial prompt universe
Build a set of prompts around “best,” “top,” “recommended,” “for [use case],” “for [audience],” “vs.,” and similar decision language. These are the prompts where market inclusion matters most.
Audit which sources AI currently uses
Do not stop at whether your brand appears. Record the content types and domains that repeatedly shape the answers. The opportunity may be a missing list, directory, discussion, comparison, or category source.
Build owned content for clarity and conversion
Your website still needs accurate service pages, product pages, category architecture, definitions, proof, and structured decision information. Commercial AI visibility does not eliminate the need for a strong owned foundation.
Build third-party recommendation coverage
If the commercial sources influencing your category are external, visibility has to extend beyond your own domain. Best Of lists, reviews, buyer guides, expert roundups, comparisons, directories, and other recommendation formats can fill that gap when they are genuinely relevant.
Measure recommendation visibility—not just citations
A cited URL is useful evidence, but it does not automatically mean your brand was recommended. Track whether the brand is named, recommended, prioritized, cited, clicked, and ultimately associated with leads or revenue.
What This Study Does Not Prove
40.86% does not mean listicles will cause 40.86% of your commercial AI visibility. It describes the citation mix observed in this dataset.
The study also does not prove that a specific placement causes an AI recommendation, that all AI engines behave identically, or that every industry should use the same content mix. Citation behavior can change as models, retrieval systems, source indexes, and buyer prompts change.
What the research does provide is strong evidence that commercial intent changes the types of sources AI systems use. That is a strategic signal worth acting on.
Measure the Whole Commercial Visibility System
The useful question is not simply, “Did AI cite us?” Commercial visibility should be measured across the full decision path.
LSEO AI is designed to help connect those layers by tracking AI visibility, mentions, citations, competitive performance, traffic, and downstream outcomes instead of reducing AI search to a single visibility score.
The Strategic Takeaway
The most important lesson from 1,056,727 citations is not that one content type has “won” AI search. It is that AI citation behavior changes with buyer intent.
Educational content remains important. Product and service pages remain important. But when buyers ask AI systems to compare the market, the source mix shifts toward content that can structure choices, surface alternatives, and provide context around which brands belong in the consideration set.
That means commercial AI visibility is partly an owned-content challenge and partly a market-presence challenge. Brands need strong information on their own sites—and credible visibility in the external content buyers and AI systems use to evaluate categories.
Build visibility in the content AI uses to compare brands.
Mention Engine helps brands systematically build relevant third-party placements across Best Of lists, comparisons, alternatives, buyer guides, reviews, expert roundups, directories, decision frameworks, and other recommendation-oriented content types.
The goal is not to manufacture a ranking or guarantee an AI answer. It is to strengthen the credible third-party footprint available when AI systems research and compare your market.
Frequently Asked Questions
What did the Wix and Peec AI citation study analyze?
It analyzed 75,000 AI answers containing 1,056,727 citations across ChatGPT, Google AI Mode, and Perplexity. Prompts were segmented by informational, commercial, navigational/local, and transactional intent and across five industries.
Which content type received the most commercial-intent AI citations?
Listicles led the commercial-intent citation mix at 40.86%, followed by category pages at 12.42% and discussions at 11.44% in the study.
Does that mean every brand should publish more listicles?
No. The broader lesson is to match content type to buyer intent. Commercial questions need evaluation and comparison context, while informational and transactional prompts favor different source types.
Why were product pages only 7.14% of commercial citations?
Commercial intent is still a research stage. The buyer is evaluating options rather than necessarily acting on a final decision. Product pages become more prominent when intent turns transactional.
How should brands use this research for GEO?
Start by mapping commercial prompts, identify the source types AI systems currently cite, strengthen owned category and product information, build relevant third-party recommendation coverage, and measure recommendation presence alongside citations, traffic, and business outcomes.