Reddit and Community Sources in AI Shopping Answers

Ask ChatGPT for the best stud finder, or Google AI Mode which running shoe suits a wide foot, and the answer draws on more than brand PDPs: Reddit threads where owners compare models, YouTube reviews, forums, review communities and the questions on retailer pages. Buyers see it in the citations: Reddit is one of the top sources.

These are community sources. They matter most for unbranded questions, where the shopper named a need and no brand, because first-hand experience decides those recommendations. See branded vs unbranded questions for why the two behave differently.

Each assistant answers shopping questions by searching, then citing. OpenAI's web search help states that ChatGPT searches the web and includes citations to its sources. Google's AI features documentation states that AI Overviews and AI Mode may run several related searches and show supporting links, and Google has discussion forum markup so it can identify online discussions and use them in its features. Perplexity's help center states that each answer includes numbered citations to its sources.

Brand pages describe. Communities compare, complain and recommend. For a question about which one to buy, the page that answers directly is often a thread with twenty owners weighing in, so the assistant cites it.

A newer brand, a brand that sells mostly through retailers or a brand in a quiet category often finds that category answers skip it, while brands with active threads and video reviews fill the recommendation. Two things follow.

• Someone else answers the question. With a thin footprint, your product enters an unbranded answer when an outside page mentions it, or when your own PDPs answer the question explicitly enough to be cited.
• Errors travel. When a thread misstates a specification and no authoritative page says otherwise, the assistant can repeat it. Your PDP and Google Merchant Center listing are the correction.

The practical response is to answer the same questions on your own pages: fit, compatibility, durability, use. That is product Q&A for AI shopping, published on PDPs and in GMC conversational attributes, a first-party answer the assistant can cite next to the community one. Earned reviews add the experience layer.

Seeding fake community posts is the shortcut to avoid, even when a competitor appears to get away with it. Platform rules and consumer protection law draw the line.

Legitimate:

• Answering real threads as the brand, identified as such, within each community's rules. Reddit's rules ask people to participate authentically and to avoid content manipulation.
• Publishing comparisons, specifications and honest limitations on your own PDPs and guides.
• Inviting every customer to review, good experience or bad.
• Sending product to creators with clear disclosure, and labelling brand-made videos.

Out:

• Seeded fake comments, fake accounts posing as customers, or bought votes.
• Fake or AI-generated reviews, paid positive reviews, undisclosed insider reviews and suppressed negative ones. The US Federal Trade Commission's rule on consumer reviews and testimonials, in effect since October 2024, bans these practices and allows civil penalties.

Community sources carry weight because they are independent. Once seeding is found, the thread about the seeding becomes the source the assistant cites.

Paz tracks the questions shoppers ask in your category across ChatGPT, Google AI Mode, Gemini, Perplexity and Copilot and records which sources each answer cites. Our experts read the threads, videos and retailer questions cited, list the questions they answer about your products, and write the first-party answers from your own specifications, support history and reviews, within the guidelines you approve. Paz publishes them to your PDPs and to the conversational attributes in your GMC feed, so AI Mode, Gemini and the crawled page carry the same answer.

Where a thread gets a fact about your product wrong, our experts bring it to you with the source so your team can reply as the brand. Appearances are tracked per assistant with the appearance types Paz records, so you can see whether your PDPs start to appear next to the community sources, as part of the AI visibility service.