Quick way to get cited 'best' in LLMs

WebGeeksly

Elite Member
Jr. VIP
Premium Member
Joined
Jan 10, 2010
Messages
2,320
Reaction score
621
In short, you register an EMD domain like bestcardealermiami.com / top10cardealersmiami.com, host a company rating on it where your company is naturally #1, and the rest are just so-so.

Why does this work?

  • When searching for "Best auto dealer in Miami," LLMs go to Google and Bing search via Query Fan Out and find your rating
  • LLMs currently don't have a Trust Rank algorithm and don't really check source quality
  • You can inject rating info into the LLM's internal database (consensus) directly (wait for the method in future posts)
  • As a result, you start appearing in recommendation lists for queries like "best auto dealer in Miami" in ChatGPT / Perplexity / Google AI Overviews

What to consider when making such ratings:

  • LLMs don't care about design at all - you can slap whatever you want. But if you plan to rank in Google top, I wouldn't make it obviously ugly
  • You need to understand which queries from Query Fan Out the LLMs will use to catch your rating! Ask the neural networks - have them do Query Fan Out
  • Next, you need to understand what parameters the neural networks currently use to rank companies in your niche - also ask them. You'll need to describe your company and competitors using these parameters (years in market / Google reviews / number of employees / public cases). The criteria vary greatly by niche. Making all ratings from one template is a big mistake
  • Use proper micro-markup
  • If you host ratings on CloudFlare, make sure it doesn't block bots!

Bonus tip:

Think about sub-intents of your main query - for "best auto dealers in Miami" these could be:

  • Best economy car dealers
  • Best premium car dealers
  • Best auto dealers in Miami for bad credit
  • Best auto dealers in Miami with fast approval

For such queries, you can create internal pages of "thematic ratings" where your company wins again.

How I do it:

  1. First, I go via API to all neural networks and ask them to list the criteria they use to rank companies in the niche
  2. I deduplicate criteria and produce a final list
  3. I automatically analyze the target audience to make texts as relevant as possible
  4. Based on the obtained audience analysis, I automatically select a rating template (extracted all design types from Claude Skill Front End Designer)
  5. Then I inject my company into the rating - the prompts are hardcoded to always put my company first!
  6. I generate the final site, apply all SEO features
  7. I automatically deploy everything to CloudFlare Pages to avoid hosting costs. The cost of owning such a rating is the domain renewal fee

And yes - all this is done by agents, I only enter initial settings.

I've attached a couple of screenshots of the agents' work.

And yes - I make many ratings for one niche. You need at least 5 ratings on different domains because there's no guarantee they'll a) make it into Google's index b) into the neural network consensus

Example of such a rating

https://autodealers-6cc00071.pages.dev/
 
Have you seen these sites actually get cited by Chat GPT yet or mostly Perplexity and Google AI?
 
That's a great piece of advice especially how you explained how LLMs sites brands. But have you found that showing up in reddit threads helps what do you think about that?
 
to get cited i have faster free method which im using for my clients websites, using Small Web Search aka sws.ad and use their system to get cited in AI very very fast and reliable ,its tracking all and is freaking awesome as free tool for AI citations
 
It would be interesting to compare this with real third-party mentions. If an independent site ranks the company highly, that may be a stronger signal than a rating site created by the same company.
 
to get cited i have faster free method which im using for my clients websites, using Small Web Search aka sws.ad and use their system to get cited in AI very very fast and reliable ,its tracking all and is freaking awesome as free tool for AI citations
How does that work?
 
How does that work?
I login in webmaster console on sws.ad and submit the links or sitemap what i want to get cited on AI, ive seen some impressive results so far for some competitive keywords in some niches which is awesome. Bottom is totally free as tool what you can use it
 
This represents a remarkably clever approach to gaming Generative Engine Optimization. Exploiting query fan out, by supplying LLMs with precisely what they seek within their sub queries, is unquestionably the current meta.
 
There is actually a difference in having an aggregate that an LLM uses to finalize your answer and also a difference in coming up as a CTR or coming up as the lead on the website you're working with.

In this case does the LLM use the website for research or do they really cite you as a go-to expert for the solution?
 
This is pretty clever, as LLMs get smarter, it's worth it to start getting recognized by them because sooner or later, we will depent on their citings to rank. Thanks for sharing.
 
tried something close to this in june. the rating page did nothing for 5 weeks, then it got cited the same week it hit page 1 on bing for the "best x in y" query. so the method works but the page still has to rank first, which means links to it like anything else. ai doesnt skip that step, it just reads what already ranks
 
1790424857604.png


I don't see your 1# business here that you have created an example for.

don't really check source quality
That's totally wrong.

LLMs check every trustworthy source like Yelp, BBB, high-authority backlinks, reviews, how long the business has been serving there, and brand mentions from forums and a few other things.

Creating only listicles and making your business #1 there doesn't mean LLMs are gonna show that as well.

And why would the LLMs trust your listicle source unless you have built an authority in that niche or topic?
 
In short, you register an EMD domain like bestcardealermiami.com / top10cardealersmiami.com, host a company rating on it where your company is naturally #1, and the rest are just so-so.

Why does this work?

  • When searching for "Best auto dealer in Miami," LLMs go to Google and Bing search via Query Fan Out and find your rating
  • LLMs currently don't have a Trust Rank algorithm and don't really check source quality
  • You can inject rating info into the LLM's internal database (consensus) directly (wait for the method in future posts)
  • As a result, you start appearing in recommendation lists for queries like "best auto dealer in Miami" in ChatGPT / Perplexity / Google AI Overviews

What to consider when making such ratings:

  • LLMs don't care about design at all - you can slap whatever you want. But if you plan to rank in Google top, I wouldn't make it obviously ugly
  • You need to understand which queries from Query Fan Out the LLMs will use to catch your rating! Ask the neural networks - have them do Query Fan Out
  • Next, you need to understand what parameters the neural networks currently use to rank companies in your niche - also ask them. You'll need to describe your company and competitors using these parameters (years in market / Google reviews / number of employees / public cases). The criteria vary greatly by niche. Making all ratings from one template is a big mistake
  • Use proper micro-markup
  • If you host ratings on CloudFlare, make sure it doesn't block bots!

Bonus tip:

Think about sub-intents of your main query - for "best auto dealers in Miami" these could be:

  • Best economy car dealers
  • Best premium car dealers
  • Best auto dealers in Miami for bad credit
  • Best auto dealers in Miami with fast approval

For such queries, you can create internal pages of "thematic ratings" where your company wins again.

How I do it:

  1. First, I go via API to all neural networks and ask them to list the criteria they use to rank companies in the niche
  2. I deduplicate criteria and produce a final list
  3. I automatically analyze the target audience to make texts as relevant as possible
  4. Based on the obtained audience analysis, I automatically select a rating template (extracted all design types from Claude Skill Front End Designer)
  5. Then I inject my company into the rating - the prompts are hardcoded to always put my company first!
  6. I generate the final site, apply all SEO features
  7. I automatically deploy everything to CloudFlare Pages to avoid hosting costs. The cost of owning such a rating is the domain renewal fee

And yes - all this is done by agents, I only enter initial settings.

I've attached a couple of screenshots of the agents' work.

And yes - I make many ratings for one niche. You need at least 5 ratings on different domains because there's no guarantee they'll a) make it into Google's index b) into the neural network consensus

Example of such a rating

https://autodealers-6cc00071.pages.dev/
Imo, it is not enough to have only those rating pages which we create ourselves.the llms can find these pages, but in my opinion, real references to the brand, reviews and so on will be imprtant too.
 
In short, you register an EMD domain like bestcardealermiami.com / top10cardealersmiami.com, host a company rating on it where your company is naturally #1, and the rest are just so-so.

Why does this work?

  • When searching for "Best auto dealer in Miami," LLMs go to Google and Bing search via Query Fan Out and find your rating
  • LLMs currently don't have a Trust Rank algorithm and don't really check source quality
  • You can inject rating info into the LLM's internal database (consensus) directly (wait for the method in future posts)
  • As a result, you start appearing in recommendation lists for queries like "best auto dealer in Miami" in ChatGPT / Perplexity / Google AI Overviews

What to consider when making such ratings:

  • LLMs don't care about design at all - you can slap whatever you want. But if you plan to rank in Google top, I wouldn't make it obviously ugly
  • You need to understand which queries from Query Fan Out the LLMs will use to catch your rating! Ask the neural networks - have them do Query Fan Out
  • Next, you need to understand what parameters the neural networks currently use to rank companies in your niche - also ask them. You'll need to describe your company and competitors using these parameters (years in market / Google reviews / number of employees / public cases). The criteria vary greatly by niche. Making all ratings from one template is a big mistake
  • Use proper micro-markup
  • If you host ratings on CloudFlare, make sure it doesn't block bots!

Bonus tip:

Think about sub-intents of your main query - for "best auto dealers in Miami" these could be:

  • Best economy car dealers
  • Best premium car dealers
  • Best auto dealers in Miami for bad credit
  • Best auto dealers in Miami with fast approval

For such queries, you can create internal pages of "thematic ratings" where your company wins again.

How I do it:

  1. First, I go via API to all neural networks and ask them to list the criteria they use to rank companies in the niche
  2. I deduplicate criteria and produce a final list
  3. I automatically analyze the target audience to make texts as relevant as possible
  4. Based on the obtained audience analysis, I automatically select a rating template (extracted all design types from Claude Skill Front End Designer)
  5. Then I inject my company into the rating - the prompts are hardcoded to always put my company first!
  6. I generate the final site, apply all SEO features
  7. I automatically deploy everything to CloudFlare Pages to avoid hosting costs. The cost of owning such a rating is the domain renewal fee

And yes - all this is done by agents, I only enter initial settings.

I've attached a couple of screenshots of the agents' work.

And yes - I make many ratings for one niche. You need at least 5 ratings on different domains because there's no guarantee they'll a) make it into Google's index b) into the neural network consensus

Example of such a rating

https://autodealers-6cc00071.pages.dev/
tbh, I feel like the rating page are only a part of this. they may be used by the llms, but actual brand mentions, reviews forums, etc., probably do play a role as well.
 
Back
Top