To outrank a competitor that AI search engines recommend instead of you, find the specific sources AI cites when it names them, then out-document those exact sources with content that is fresher, more specific, and corroborated by more independent voices. That means identifying the Reddit threads, review pages, "best of" roundups, and comparison articles AI quotes for the competitor, then earning your own accurate, current presence on each one. Around 85% of AI citations point to third-party sites, so displacement is won on those pages, not on your own blog. This is the work of AEO. This guide gives the steps.
You are not fighting a ranking algorithm here. You are fighting a paper trail. The competitor wins because a handful of pages say good, specific things about them and either say nothing about you or say something stale. Change the paper trail and you change the answer.
How do you find which sources AI cites for a competitor?
Ask the AI search engine the buying question your customers ask, name no brand, and read the citations it attaches to the competitor it recommends. ChatGPT, Perplexity, Gemini, Grok, and Claude all surface or footnote the URLs they pulled from, so the recommendation comes with a visible bibliography. That bibliography is your target list.
Run the category and use-case prompts your buyers actually type, such as "best [category] for [specific use case]" or "[competitor] vs alternatives," across every engine you can. When a competitor appears at position one, expand the citations and write down every domain: the Reddit thread, the G2 or Capterra page, the roundup article, the competitor's own comparison page. Repeat the same prompts a few times, because AI search results change between runs and you want the sources that recur, not the one-off ones.
What to do: Build a spreadsheet with one row per cited source. Columns: the URL, which engine cited it, what it says about the competitor, and whether it mentions you at all. The recurring URLs across multiple engines are where you start. Those are the load-bearing pages holding up the competitor's recommendation.
Why displacing the source beats improving your own website
You displace a competitor by changing what third-party sources say, because AI search engines weigh independent coverage far more heavily than anything on your own domain. Roughly 85% of citations point to sites other than the brand being asked about, so a perfect homepage does almost nothing if the Reddit thread and the review page still favor the competitor. The competitor's recommendation lives on pages you do not own, which means your fix lives there too.
This is the part that frustrates founders who have just rebuilt their site. You can have better software, better service, and a better price, and still lose the answer, because AI is reading a two-year-old roundup that named your competitor and never heard of you. The roundup is the problem. Your site is not. For the full reasoning on why this happens, see why AI citations come from third-party sites.
What to do: Treat each cited source as a separate campaign. For every URL in your spreadsheet, decide whether you can get added to it, get a fresher competing page to outrank it, or earn a new citation alongside it. You are not trying to delete the competitor. You are trying to make the next page AI reads include you, with sharper specifics.
Out-document the Reddit threads AI quotes
When AI cites a Reddit thread that recommends a competitor, the move is to surface a more recent, more specific, and more corroborated answer in the same subreddit, not to bury the old thread. Grok and Perplexity lean heavily on Reddit, so a single popular thread can carry a competitor across multiple engines. A genuine, detailed comment or a fresh thread that answers the exact buying question with concrete specifics gives AI a newer source to quote.
Find the threads from your citation list, read why the competitor got named (usually a specific use case, a price point, or a feature), and answer that same need with more detail than the original. If the thread says "Competitor X is great for small teams," and you genuinely serve small teams better, say exactly how: the seat price, the onboarding time, the one feature that matters. Specificity is what AI extracts.
What to do: Participate as a real account in the subreddits your buyers use, answer the actual question asked, and include the concrete numbers AI loves to quote: pricing, timelines, and named use cases. Never astroturf. Fake or thin posts get removed, and a removed thread cites nobody. For the durable version of this play, see how to build third-party presence for AI search.
Get into the "best of" roundups that rank for your category
To displace a competitor in the roundups AI cites, get yourself added to the existing high-ranking lists and, where you cannot, publish a fresher comparison that out-ranks the stale one. AI search engines treat "best [category]" listicles as editorial fact and quote whoever appears in them, often regardless of how current the list is. If the top three roundups for your category name the competitor and omit you, those three pages are doing the competitor's selling.
Identify the roundup authors from your citation list and pitch them with a specific, verifiable reason to include you: a feature the competitor lacks, a price advantage, a use case you serve better. Editors update lists when given a concrete, checkable claim, far more readily than when given a generic "please add us." Where a list is abandoned, the counter-move is to publish your own comparison that covers the full landscape, names competitors honestly, includes pricing, and stays current, because AI favors recent content and will start quoting the fresher page.
What to do: For each roundup, send a one-paragraph inclusion pitch with one falsifiable differentiator and a link to proof. For categories where the lists are stale, write the comparison article you wish existed, update it monthly, and let recency do the displacing.
Counter the competitor's own comparison pages
When AI cites a competitor's "[Competitor] vs You" or "alternatives to [Competitor]" page, publish your own comparison that tells the full story with current pricing and honest trade-offs, because AI rewards comprehensive coverage over one-sided marketing. Competitors write these pages to frame the matchup on their terms, and if yours is the only voice missing, AI inherits their framing. The fix is to enter the comparison conversation with a page AI can quote instead.
A comparison page that names both products, lists real pricing, and concedes where the competitor is genuinely better reads to AI like editorial coverage rather than an ad, which is exactly the kind of source it cites. A page that only says you are better looks like marketing and gets ignored. The goal is the most useful, most specific page on the matchup, not the most flattering one.
What to do: Publish a comparison page for each competitor that AI cites, covering pricing by tier, what each tool includes, who each is built for, and where the competitor genuinely wins. Keep it updated so it stays inside AI's recency window. Honest, current, and specific beats glowing and stale.
Make every claim more specific and more corroborated
The deciding factor between two candidates is which one is described more specifically and confirmed by more independent sources, so win on both. AI search engines do not just count mentions; they read what each source says and build a narrative about which brand fits the user's exact intent. A competitor described as "good for protein-heavy prep, high-carbon steel, holds an edge" beats a brand described as "a solid all-rounder," even if the all-rounder is objectively better, because the first description answers a specific intent.
Specificity wins the read; corroboration wins the trust. If three independent pages say the same specific thing about you, AI treats it as fact. If only your website says it, AI treats it as a claim. The work is to get the same precise, true statement about your strengths onto several sources AI already reads, so they reinforce each other. This is the same mechanic behind why ChatGPT recommends your competitors in the first place.
What to do: Write one sharp, true sentence about what you are best at for a specific use case, then get that exact claim, backed by a number, onto your site, a review platform, a Reddit answer, and a roundup. Repeat for each use case you want to win. Vague and singular loses to specific and corroborated every time.
A 30-day displacement plan
Displacing a competitor in AI search is a four-week sequence: map the sources, prioritize them, out-document the highest-impact ones, then verify the answer changed. None of these steps touches a ranking dial, because there isn't one. They change the sources AI reads, which changes what it says.
- Week 1, map. Run your buyers' category and use-case prompts across every AI search engine, record every source cited for the competitor, and flag the URLs that recur across engines.
- Week 2, prioritize. Rank the sources by how often they are cited and how easy they are to influence. A widely cited Reddit thread you can answer beats a single editorial mention you cannot touch.
- Week 3, out-document. Work top down. Answer the Reddit threads with specifics, pitch the roundups with falsifiable differentiators, publish your comparison pages, and push one corroborated claim across multiple sources.
- Week 4, verify. Re-run the original prompts and check whether you now appear, where you rank, and which of your new sources AI quotes. Track this monthly, because a single check measures noise. For a deeper diagnosis of the gap, see competitors in ChatGPT but not me.
Loudmink is an AEO platform that maps the exact sources AI search engines cite for you and your competitors, then creates content across blog, Reddit, and YouTube to out-document them, with post-publication verification that rechecks the engines. Plans from $99/mo as of June 2026. Start with a free scan, or track the gap continuously with visibility tracking.
Frequently Asked Questions
How do I find out which sources AI recommends my competitor from?
Ask each AI search engine the buying question your customers ask without naming any brand, then read the citations it attaches to the competitor it recommends. ChatGPT, Perplexity, Gemini, Grok, and Claude show or footnote the URLs they used, so the recommendation comes with a source list. The URLs that recur across multiple engines and multiple runs are the ones holding up the recommendation.
Can I really outrank a competitor that AI already recommends?
Yes, because AI recommendations are built from third-party sources you can influence, not a fixed ranking. Around 85% of AI citations point to sites other than the brand being asked about, so getting added to the roundups, review pages, and Reddit threads AI cites for the competitor, with fresher and more specific content, changes the answer over weeks to months.
Should I attack the competitor or improve my own coverage?
Improve your own coverage on the sources AI reads. You cannot remove a competitor's mentions, and trying to looks like astroturf, which gets content deleted. The durable play is to make the next page AI reads include you with sharper, corroborated specifics, so your name enters the same conversation rather than trying to erase theirs.
How long does it take to displace a competitor in AI search?
Plan for weeks to a few months. New third-party content needs to be indexed, picked up by AI's live search, and corroborated across enough sources to shift the narrative. AI also favors recent content, so keeping your comparison pages and roundup placements current is what holds the gain once you make it.