Original article by Leigh McKenzie and Shannon Willoby
30-second rundown:
- DO build one piece of LLM-ready content this week: pick a “best of” topic in your niche, explain your selection criteria upfront, give each item a specific “best for” label, and structure every entry identically with a summary, key features, pros, cons, and pricing.
- DO open Reddit right now, find the two or three subreddits where your audience actually lives, and start answering questions with genuine expertise. No pitch. No links. Just show up consistently, because Reddit is the single most-cited source by LLMs and most of your competitors haven’t figured that out yet.
- DON’T wait until next quarter to check if any of this is working. Open Google Analytics today, pull your direct traffic for the last six months, and compare it against your organic clicks in Google Search Console. If direct is climbing while organic clicks are falling, LLMs may already be mentioning you. If both are falling, you are invisible to the machines and the clock is running.
Leigh McKenzie and Shannon Willoby wrote this dispatch for Backlinko, and it opens with a gut-punch that deserves to be stitched on a banner above every marketing department in the world: Backlinko doesn’t rank number one for “best SEO resources,” but ChatGPT mentioned them twice anyway.
That is not luck.
That is a new kind of war, and most people are still lacing up their boots while the bombs are already falling.
1. Citations Are the New Rankings.
LLM seeding is the deliberate practice of publishing content in the formats and locations that large language models scrape, summarize, and cite.
You are no longer optimizing for clicks. You are optimizing for citations.
You are no longer building backlinks. You are engineering brand mentions.
And here is the statistic that should rearrange your priorities permanently: according to a Semrush study cited in the article, almost 90% of ChatGPT citations come from pages sitting at position 21 or beyond in Google.
The guy with the fat, comfortable number-one ranking is getting ignored by the machine while your page-four comparison post gets quoted to ten million strangers.
The playing field did not level.
It inverted.
2. Format Is the Weapon, Structure Is the Ammunition.
The article details exactly what content LLMs feast on, and none of it is mysterious.
Structured “best of” lists that explain upfront how items were selected. First-person product reviews with measurable outcomes, repeatable testing processes, and honest tradeoffs between strengths and weaknesses.
Comparison tables that go beyond features to deliver use-case verdicts: who this is for, who it isn’t, and why. FAQ sections written as direct subheadings with answers that start on the first line, not the fourth. Free tools, templates, and frameworks with plain-language descriptions that explain who they’re for and how to use them.
The connective tissue across all of it is what the article calls semantic chunking: short, clearly labeled sections, each focused on a single idea, with a consistent repeatable format for every entry.
WordPress plugins like RankMath and Yoast can even add FAQPage structured data automatically to accelerate LLM pickup.
The machines are not subtle.
Feed them clean, well-organized, honest content and they will remember your name.
3. Where You Plant Matters as Much as What You Plant.
The article maps the watering holes with the precision of a field guide for a territory most marketers haven’t scouted yet.
- Reddit is the single most-cited source by LLMs, according to Semrush. Show up in relevant subreddits with genuine expertise and no sales pitch, and do it consistently.
- Quora is the most cited website specifically in Google’s AI Overviews, and it rewards the same approach: comprehensive answers, clear formatting, specific examples.
- Medium and Substack carry structural credibility that makes them easy for AI to parse.
- LinkedIn articles are indexed well and tied to real professional identities, which gives them a credibility bonus.
- Third-party review platforms like G2, Capterra, and TrustRadius are LLM magnets because they combine feature breakdowns, user reviews, and pros-and-cons in a format the machines find irresistible.
Encourage your customers to leave detailed, specific reviews and prompt them with real questions about outcomes and comparisons.
4. Build a Base the Machines Already Trust.
For brands with the stomach for it, the article also recommends editorial-style microsites structured like independent publications rather than company pages, using IKEA’s Life at Home research site as a specimen. The lesson is the same everywhere you look: get your name into the places the machines trust, before your competitors figure out the map.
The article’s closing argument is the most useful and the most ignored part.
5. The Instruments You Need for an Invisible War.
Track your branded and direct traffic in Google Analytics. If direct traffic is rising while organic clicks are falling, the LLMs are quite possibly working.
Run manual prompts across ChatGPT, Claude, Perplexity, and Gemini in incognito, searching the way your audience searches, and document every mention with the tool used, the exact language, and where you placed in the response.
Set alerts for your brand name in Semrush Brand Monitoring, SparkToro, or plain old Google Alerts.
The war is invisible if you’re not watching, but it is very much underway.
