# Get on the best of lists AI cites: roundups, Reddit, YouTube > In short: Find the third-party pages the assistants cite for your buying questions, then earn a place on each one the way a good PR person would: a correction where they have your facts wrong, a factual one-page sheet and a free account for a roundup's author, honest participation in the threads your buyers read, and reviews from real customers. Don't pay for placement and don't astroturf. Guide from Keats, published 2026-11-13, updated 2026-11-13. Canonical: https://getkeats.com/guides/third-party-placements On a question like "which exam prep course is best for working adults?", the page that decides the answer often isn't anyone's own site. It's a roundup on a well-ranked publication, a Reddit thread, a YouTube review, a comparison on a review site. If three rivals are on that page and you're not, the assistant has nothing to name you from. This is the slowest part of AEO, and on category questions it's often the part that matters most. ## How much third-party pages count You can see it in the citations. On the six AEO searches we checked on Google, AI Overviews cited YouTube on five and Reddit on four, alongside LinkedIn posts and publishers' articles ([the six searches](https://getkeats.com/data/google-serps-2026-11-12.csv)). When we asked ChatGPT for AI visibility tools for a small marketing team, most of the six pages it cited were roundups ([the sample](https://getkeats.com/data/chatgpt-sample-2026-11-12.csv)). Reddit is also wired in at the source. Google's 2024 announcement of its expanded partnership with Reddit says the Reddit Data API gives Google "efficient and structured access to fresher information" from the site. Whatever the exact weight, community discussion is part of what assistants read. ## Step 1: build the list of pages that matter For each buying question you track, collect every third-party page the assistants cite, across ChatGPT, Gemini, Perplexity and Google's AI features. Then sort: - **Pages that name rivals and not you.** Your target list. - **Pages that name you wrongly.** Old price, wrong category, a dropped product. Fix these first; they're cheap and they're already being read. - **Pages that name you.** Keep them accurate, and note who wrote them. You'll usually find the same few domains again and again. Ten pages tend to carry most of the weight for a category. ## Step 2: roundups and listicles The author of a "best X for Y" article wants it to be accurate and current, because readers complain when it isn't. Make it easy: - **Send a one-page fact sheet**: what you sell, who it's for, prices, what makes you different, with links. No adjectives. - **Offer a free account** so they can try it themselves. - **Fix their facts politely** if you're already listed wrongly. - **Suggest where you fit**, honestly: "for working adults" is more likely to be added than "the best overall". Don't pay for placement. Paid inclusion tends to be disclosed or found out, and when a buyer discovers it, the placement costs more trust than it bought. ## Step 3: Reddit and communities Assistants cite threads where real people compare options. You can't and shouldn't manufacture those, but you can be present in them: - **Answer questions in your field** where you know the answer, as a named person who says where they work. - **Disclose every time** you mention your own product. - **Don't post with fake accounts, buy upvotes or have an agent post for you.** Communities spot it, moderators remove it, and a removed thread full of shills is the worst page to be cited from. - **Ask happy customers to share their experience** where it's relevant, with no script. ## Step 4: YouTube and video Video results show up repeatedly in AI Overviews. The videos that get cited tend to be plain: a walkthrough, a comparison, a "how I did X". A clear title that matches the question, a description that states the answer, and an accurate transcript (YouTube's captions, checked) give a model something to read. ## Step 5: review sites Review sites are cited for buying questions in most software and consumer categories. Keep your profiles complete and current: category, pricing, features, screenshots. Ask real customers for reviews at a natural moment, such as after a success, and reply to the critical ones factually. Don't gate, incentivise or write reviews; most platforms forbid it, and it's the kind of thing that gets profiles flagged. ## How long it takes Weeks to months. An author may update a roundup quarterly, and an assistant picks up the change when the page is re-crawled. That's why [getting named in ChatGPT](https://getkeats.com/guides/get-named-in-chatgpt) puts it last: fix your own pages first, then work outward. Track the third-party pages you've worked on as a change like any other, and measure whether the questions they serve moved. For Perplexity, which shows every source next to its answer, the list is easiest to build; our guide to [Perplexity citations](https://getkeats.com/guides/perplexity-citations) covers it. And if you're starting from "why do they keep naming my competitor?", start with [that diagnosis](https://getkeats.com/guides/why-chatgpt-recommends-your-competitor), which tells you which pages to go after. ## Your agent does this Ask your agent: "List the third-party pages cited on our buying questions that name rivals and not us, grouped by site, with the author or contact for each." It reads `list_citations` and `brands_named` and drafts a short, factual note for each, for you to send. ## Sources - [Google: An expanded partnership with Reddit](https://blog.google/company-news/inside-google/company-announcements/expanded-reddit-partnership/), read 2026-11-13 - [Keats: Google results and AI Overview citations for six AEO searches (CSV)](https://getkeats.com/data/google-serps-2026-11-12.csv), read 2026-11-13 - [Keats: eight ChatGPT answers to buyers' questions about AEO tools (CSV)](https://getkeats.com/data/chatgpt-sample-2026-11-12.csv), read 2026-11-13