Fable 5 Masterclass for LLM SEO (full B2B playbook)

we've spent the last 2 years running LLM SEO for over 100 B2B companies...

SaaS platforms, agencies, ecommerce brands, and service businesses across dozens of categories.

we've seen what actually drives citations across ChatGPT, Claude, Perplexity, and Gemini.

we've also seen the mountain of audit work that stops most founders from ever building an LLM SEO strategy in the first place.

Fable 5 just dropped.

we've built out a 6-prompt system inside it that runs the LLM SEO audit work we used to spend entire weeks on.

full audit of where your brand is invisible in AI answers.

full map of what URLs your competitors are winning citation slots with.

full comparison-page gap list.

AI Overview trigger map.

Reddit citation trail.

entity association audit.

every prompt below is written out end-to-end.

the exact wording. what to plug in. what output to expect. and why each one moves the needle for AI citations.

bookmark it. you'll come back to this.

1. AI category visibility audit

Google rankings and AI citations aren't the same game anymore.

your brand can sit at position 3 on Google for a category term and be completely absent when a buyer asks ChatGPT the same question.

Omni Eclipse ran a 1,700-business audit earlier this year.

88% of businesses are invisible in ChatGPT.

77% of the businesses ranking on Google's first page are also invisible in ChatGPT specifically.

the two systems have diverged.

AI engines pull citations from Reddit, LinkedIn, Wikipedia, YouTube, and comparison pages way more heavily than your homepage.

so the only way to know where you actually stand is to test each engine directly with the queries your buyers are typing.

the prompt:

Run the following 20 buyer queries across ChatGPT, Perplexity, Claude, and Google AI Overviews. Queries: [list your 20 highest-intent category queries here]. For each query on each engine, log: which brands are cited, in what position, what URL each brand is cited from, and any brand recommended as the #1 answer. Output a spreadsheet. Rows: queries. Columns: each AI engine. Cells: which brands are cited and their positions. Add a summary tab that shows my brand's total citation count across all 20 queries per engine.

why this matters:

the spreadsheet you get back is your baseline.

you'll see exactly where you're being cited, where you're not, and which AI engines are your worst blind spots.

a lot of the time founders discover they're being cited by Perplexity but completely invisible in ChatGPT, which serves 800 million users a week.

running this manually across 20 queries × 4 engines takes 8-10 hours.

Fable 5 does it in under 30 minutes.

2. Competitor citation map

once you know where you're invisible, the next question is who's winning those citation slots and what URLs they're using to win them.

this is where most LLM SEO breaks down.

founders try to reverse-engineer competitor rankings by looking at their homepage or their landing pages.

but the URL that's actually driving citations is often a comparison page, a Reddit thread, a G2 profile, or a LinkedIn post their founder wrote 6 months ago.

76% of AI Overview citations come from URLs already ranking in Google's top 10.

each engine weights different sources on top of that.

ChatGPT pulls 47.9% of its citations from Wikipedia.

Perplexity pulls 46.7% from Reddit.

Google AI Overviews pull heavily from YouTube and Quora.

the prompt:

Run the same 20 buyer queries across ChatGPT, Perplexity, Claude, and Google AI Overviews. For every citation returned, extract the source URL, the domain, and the type of page (homepage, landing page, comparison page, Reddit thread, YouTube video, LinkedIn post, G2 profile, etc). Build a spreadsheet with one row per citation. Columns: query, engine, cited brand, source URL, source domain, page type. Then add a summary tab showing the top 20 domains driving citations across all queries, ranked by citation count.

why this matters:

you now know exactly which types of URLs are getting cited in your category.

if 40% of your competitors' citations come from Reddit threads, you know where to focus your effort.

if 30% come from G2 comparison pages, that's your next play.

no more guessing.

3. Comparison page gap analysis

comparison and alternatives pages get cited more than any other content type for category recommendation queries.

when a buyer asks ChatGPT "best CRM for remote teams" or "alternatives to Asana," the engine grabs the most authoritative comparison page indexed on Google.

if you don't have that comparison page, your competitors do.

and their brand is what gets recommended.

the mechanic worth understanding: AI engines don't pull comparison pages randomly.

they pull the ones with the strongest topical authority + freshness signals.

a comparison page updated 2 months ago outranks a comparison page updated 2 years ago in almost every case.

the prompt:

Crawl my domain [your URL] and 3 competitor domains [competitor URLs]. For each site, list every comparison page (X vs Y), alternatives page (alternatives to X), and category listicle (best X for Y) they have published. Put it in a spreadsheet. Columns: page type, target keyword, my site (yes/no), competitor 1 (yes/no), competitor 2 (yes/no), competitor 3 (yes/no), last updated date. Highlight every combo where a competitor has the page and I don't. Then for my top 10 missing pages, check which URLs currently rank in the top 5 organic results on Google. Note their title tag, first heading, and how many words each page is.

why this matters:

you now have a prioritized comparison page build list with two layers baked in.

what's missing, and what quality bar you have to hit to outrank the current winner.

no need to pay $2,000-3,000 for an outside comparison audit anymore.

we've seen B2B brands go from invisible on category queries to consistently cited across ChatGPT, Perplexity, and Gemini inside 90 days by systematically filling these gaps.

4. AI Overview trigger map

Google AI Overviews now appear on 25% of all searches, up from 6.49% in January 2025.

but they don't appear on every keyword.

when they do, position 1 CTR drops from 1.41% to 0.64% (Ahrefs 300K keyword study).

that means every one of your target keywords needs to be audited against AI Overview presence.

because if AI Overview is showing on the query and you're not cited inside it, your organic CTR falls off a cliff.

the prompt:

Run each of my target keywords [paste your keyword list] through a Google search. For each keyword, log: does an AI Overview appear yes/no, how many sources are cited in the AI Overview, which domains those sources are on, whether my domain appears in the AI Overview, and whether my top competitor's domain appears. Output a spreadsheet. Rows: keywords. Columns: AI Overview appears, my domain cited, competitor 1 cited, competitor 2 cited, competitor 3 cited, cited domains list. Sort by AI Overview appearance rate.

why this matters:

you now have a decision matrix for every target keyword.

for keywords where an AI Overview appears but your domain isn't cited, that's a priority for optimizing toward AI Overview inclusion.

for keywords where no AI Overview appears, you can still play the traditional organic game.

for keywords where AI Overviews dominate and there's no realistic inclusion path, you can save the effort and focus elsewhere.

checking every keyword manually on incognito Google is 20-30 hours for a proper list.

Fable 5 does it in a couple of hours.

5. Reddit citation analysis

Perplexity pulls 46.7% of its citations from Reddit.

Google AI Overviews pull 21% from Reddit.

ChatGPT pulls 11.3% from Reddit.

if you're serious about AI citations, you need to know exactly which subreddit threads are being pulled into AI answers for your category.

those threads are the citation surface you either need to be inside or you need to build a Reddit strategy to match.

the mechanic worth understanding: the Reddit threads getting cited aren't always the most upvoted or the oldest.

AI engines weight recent engagement and comment quality heavily.

a thread with 200 upvotes and 30 recent comments outperforms a thread with 5,000 upvotes from 3 years ago.

the prompt:

Run these 15 buyer queries in my category [paste queries] through Perplexity. For every Reddit thread cited across all queries, extract: the subreddit name, thread title, thread URL, upvote count, comment count, date posted, date of most recent comment, and which brands are mentioned in the comments. Output a spreadsheet. Rows: cited threads. Columns: subreddit, title, URL, upvotes, comments, post date, latest comment date, brands mentioned. Add a summary tab showing the top 10 subreddits driving citations for my category.

why this matters:

you now have the exact list of subreddits where buyers in your category are actively discussing solutions.

these are where your brand needs to be showing up in the conversation.

not as a promo. as a helpful contributor who actually knows the space.

we've seen brands go from zero AI citations to consistent Perplexity citations inside 60 days by targeting the exact subreddits this analysis surfaces.

6. Entity association audit

this is the most overlooked prompt on the list.

it's also the one with the highest ROI when you fix what it surfaces.

AI engines build entity associations.

when someone asks "what is X known for," the AI generates its answer based on which topics and attributes co-occur with your brand name across all the sources it was trained on.

a brand can rank #1 for their own name on Google but be associated with completely wrong topics by AI engines.

someone types "Musicfy" into Google and lands on their homepage.

but ask ChatGPT "what is Musicfy known for" and the engine might associate them with "Drake AI voice" instead of "AI music generation platform" broadly.

that's why 2 brands with equal SEO rankings can get completely different AI citation results.

the one with cleaner entity associations wins across every engine at once.

the prompt:

For each of these 5 brand names [my brand + 4 competitors], run the following 4 queries across ChatGPT, Claude, and Perplexity: What is [brand] known for? What does [brand] do? Who uses [brand]? What category is [brand] in? For each response, extract the top 5 topics/attributes the AI associates with each brand. Output a spreadsheet. Rows: brands. Columns: engine + query type. Cells: the top 5 associations. Highlight any misalignment between what the brand wants to be known for vs what the AI actually says.

why this matters:

you now know exactly where your entity signals need to be reinforced.

if AI engines think you're known for X but you want to be known for Y, you need more content, mentions, and structured data reinforcing Y across the web.

fixing entity misalignment is the highest ROI move in LLM SEO because a single clean association can unlock citation slots across every AI engine at once.

running this manually across 5 brands × 4 queries × 3 engines is 60 individual searches.

Fable 5 handles it in one workflow.

Where AI still won't help

Fable 5 handles research.

it doesn't handle judgment.

it won't tell you which of your competitors is beatable and which one owns a distribution advantage you'll never catch.

it won't know that the reason Notion gets cited for "team collaboration tools" is because they got mentioned in every productivity newsletter in 2022.

it won't design the LLM SEO strategy from scratch.

the audit, mapping, and comparison work is what Fable 5 is genuinely useful for.

the strategic layer sitting on top of that is still where an experienced operator (yours or someone you hire) adds most of the value.

How to sequence this

running all 6 in a single sitting is what most people try and burn out on.

spread them across a month so the audit findings actually inform the next round of prompts.

start with visibility (prompt 1) so you know your baseline citation count across each AI engine.

that data feeds prompt 2 (competitor citation mapping) because you now know which engines you're losing to whom.

comparison page analysis (prompt 3) and AI Overview mapping (prompt 4) come next.

these two together give you your content plan for the next quarter.

end the month with Reddit (prompt 5) and entity audit (prompt 6).

these are the off-page and positioning moves that unlock citation slots once your content foundation is set.

re-run prompt 1 every 30 days after that to measure your citation growth over time.

odds are you're going to bookmark this and never open Fable 5.

that's how these things go.

no judgement.

if you'd rather have us build the whole citation strategy for your brand instead, that's what we do.

we've driven $35M+ in client revenue getting B2B brands cited across ChatGPT, Claude, Perplexity, and Gemini using variations of the exact audit and mapping work above.

book a discovery call if you want to talk it through:

https://trailblazermktg.com/#meeting-calender

Links

• trailblazermktg.com/#meeting-calender
• x.com/search?q=%2435M%2B&src=cashtag_click
• trailblazermktg.com/#meeting-calender**