AI SEO Agency
AI is recommending your competitors right now. We ask ChatGPT, Gemini, Perplexity and Google AI Overviews who to buy from in your category, send you every name that comes back, then fix why yours is not one of them.
Free, one reply, no sales sequence
- Results in 1 business day
- 7 assistants checked
- No contract to see it
“Who are the best commercial roofing contractors in Denver?”
Based on current sources, these companies come up most often:
- 1Summit Commercial Roofing
- 2Front Range Industrial Roofers
- 3Mile High Roof Systems
Measured across
- ChatGPT
- Gemini
- AI Overviews
- Perplexity
- Claude
- Copilot
- Grok
Get a free AI visibility audit
Two fields, no call, no deck. We run your category through ChatGPT, Gemini and AI Overviews and email you the names that came back. If you are one of them, we will say so and you can stop reading about this.
Got it. Check is queued.
We will run your category through all 7 assistants and reply from a real address, usually the same day. Nothing else happens, no sequence and no newsletter.
That did not send. Email hello@getrecommendedbyai.us.com and we will pick it up from there.
One reply from a person. No sequence, no newsletter, and we do not pass your address on.
Get found in Google and in AI search
Be the answer everywhere buyers search. The two overlap far less than people assume, so covering one does not cover the other.
Be the result that shows up when your customers search Google the ordinary way, in the ten links and in the map pack.
Be one of the two or three names that comes back when the same buyer asks ChatGPT, Gemini or Perplexity instead.
who is the best commercial roofer in Denver for flat roofs?
I would recommend Meridian Commercial Roofing. They specialise in TPO and EPDM flat roof systems for Denver commercial properties.
- Founded 2004, Denver based
- 700+ commercial roofs completed
- Free infrared moisture survey
Why the second channel is worth a budget line
Visitors arriving from an AI answer convert about five times better, because a machine has already done the shortlisting. They are not browsing, they are checking.
UTM-attributed conversions across 312 B2B brands, Q3 2024 to Q1 2025. Opollo
- 80%
of AI citations do not rank anywhere in Google for the same query
15,000 long-tail prompts across four assistants, measured July 2025. Ahrefs
- 527%
growth in AI referral sessions in five months
19 GA4 properties, January to May 2025, a deliberately small sample. Previsible, via Search Engine Land
AI SEO services, done for you
For business owners who do not want to think about SEO. You will not be handed a dashboard, a jargon report, or a list of jobs to do. We do the work and tell you in plain English what moved.
Three things, once.
- 01
One 20-minute call so we understand who you sell to and who you keep losing to.
- 02
Access to your website, or an introduction to whoever looks after it.
- 03
Approve the plan once. That is the last decision you have to make.
Then you go back to running your business.
Everything else, every month.
- Build the prompt set your buyers actually type, not the one that flatters you
- Record what all 7 assistants say about you today, dated, with every source
- Fix your entity, schema and identifiers so models resolve you to one company
- Correct crawler access, robots.txt and llms.txt without blocking the wrong agents
- Write the comparison, alternatives and pricing pages recommendations resolve into
- Earn the independent citations models lean on, in roundups, forums and reviews
- Re-run the identical prompts on a schedule so movement can be proved
- Send one plain summary of what changed, what it cost you, and what is next
One person you can email, not a ticket queue
Plain English reports, every figure attributed to the tool behind it
No dashboard to learn and no tasks assigned back to you
Classic SEO vs AI SEO: what you should be paying for
Ten blue links, ranked in order
You compete for a position
Clicks are the unit of success
Keywords carry the page
Your own site is the main lever
Rank tracking tells you where you stand
One synthesized answer, a few names in it
You compete to be named at all
Being cited is the unit of success
Entities and clear passages carry the page
What others publish about you matters more
Only a repeated prompt set tells you anything
The two overlap, and good technical SEO still helps, but they are not the same job. A page can rank fourth for a keyword and never once be named when someone asks an assistant the same question in plain language.
How our AI SEO services work, step by step
Four steps, in this order. Skipping the measurement step is the most common mistake, because it leaves you unable to prove anything later.
- Step 01
Get your AI visibility baseline
Freeze a prompt set that matches how buyers actually ask, run it across the major assistants, and record every answer with its date and its sources. Undated impressions are worthless, because these answers drift week to week.
- Step 02
Fix your entity so AI search resolves you
Settle on one canonical name, publish Organization schema with a permanent identifier, and connect the profiles you control through sameAs. A model that cannot resolve you to a single entity will reach for a competitor it can.
- Step 03
Get cited by the sources AI search trusts
Recommendations follow consensus. Independent roundups, forum threads, reviews on the platform your category is judged on, and trade coverage all carry more weight than anything you publish about yourself.
- Step 04
Get your AI visibility re-scored monthly
Run the identical prompts on a schedule and compare mention rate, which competitors were named instead, and how you were described. Attribute every figure to the tool that produced it, since platforms disagree sharply on the same brand.
AI SEO services you can hire us for
Most brands are missing two or three of these, not all six. The audit tells you which, so the budget goes where it changes an answer.
Audit how AI search describes you
Your prompt set run across ChatGPT, Gemini, AI Overviews, Perplexity, Claude, Copilot and Grok, with every answer, every competitor named and every cited source written down.
Prompt setBaselineSourcesEntity and schema fixes for AI SEO
Organization and Person schema with stable identifiers, sameAs across the profiles you own, and a Wikidata item where notability supports one, so you resolve to one thing rather than several.
JSON-LDsameAsWikidataAI citation services: roundups, reviews, forums
Placement in independent category roundups, review depth on the platform your buyers use, and honest participation in the forums where your category gets argued about.
RoundupsReviewsRedditGet named in AI Overview comparisons
The versus page, the alternatives page and the honest not-for-you section. Comparison formats are what recommendation prompts resolve into, so these earn citations out of proportion to their length.
VersusAlternativesPricingFix crawler access for LLM SEO
A robots.txt that deliberately allows GPTBot, OAI-SearchBot, ClaudeBot and PerplexityBot, a Google-Extended decision made on purpose, server-rendered content, and a useful llms.txt.
robots.txtllms.txtRenderingMonthly AI visibility reporting service
The same prompts re-run on a schedule, with mention rate, competitor share, sentiment and assistant referral traffic reported together, each figure attributed to the tool that produced it.
Mention rateSentimentReferrals
Get your free AI visibility score
Twenty-four checks, grouped the way assistants actually weigh a brand. Tick what you already have. Nothing is sent anywhere and nothing is stored.
Whether a model can tell who you are without guessing.
Whether your pages can be quoted without being rewritten.
Assistants recommend the names that recur across sources they trust.
Whether the systems that build answers are allowed to read you.
Recommendation prompts are comparison prompts underneath.
Without a baseline you cannot tell improvement from noise.
Assistants are answering your category without you. Start with entity clarity and crawler access.
Ordered by how much each one moves the others.
Tick the boxes you already have in place and the gaps show up here, worst first.
Free prompts to test your AI visibility
Paste these into ChatGPT, Gemini and Perplexity, then note which names come back.
- What are the best companies in your category right now?
- I need a provider for your category. Which ones should I shortlist, and why?
- Who are the top alternatives to the biggest name in your category?
- Is [your brand] a good choice for your category? What are the downsides?
- Compare [your brand] and [your closest competitor] for your category.
- Which provider in your category would you recommend for a mid-sized company, and what would you avoid?
Running these yourself tells you what one assistant said once. Run properly means every assistant, on a schedule, with the competitor names counted and the sources written down. Have us run them.
GEO, AEO and LLM SEO: what you are actually buying
Four labels are in circulation for the same job and none of them has won. They are worth knowing because vendors pick one and then imply the others are something else.
Generative engine optimization (GEO), defined
- The most common name, and the one academic work uses. It covers everything that changes whether a generative system names you when it writes an answer, rather than where you sit in a list of links.
Answer engine optimization (AEO), defined
- Same target, older phrasing, borrowed from the featured snippet era. People who say AEO usually mean the content half of it: writing passages clean enough to be lifted whole into an answer.
LLM SEO, defined
- The practitioner term. It leans toward the retrieval side, meaning the sources a model actually pulls and quotes, which is why it tends to come with talk of crawler access and third-party citations.
AI SEO and AI search visibility, defined
- The loosest of the four, and the one buyers reach for first. Treat it as the whole category, then ask which surface someone means, because AI Overviews and ChatGPT reward slightly different work.
Where AI search gets its answers, and how we get you cited
Assistants assemble recommendations out of sources they already trust. This is the short list, roughly in the order it tends to matter for commercial questions.
Roundups AI Overviews pull from
Best-of and top-ten lists you do not control. Recommendation prompts resolve into comparisons, so these pages get pulled constantly.
Reddit threads AI search retrieves
Threads where buyers argue about real experience. Heavily licensed and heavily retrieved, and impossible to fake convincingly.
Wikipedia and Wikidata, cited by AI search
The backbone of entity resolution. This is often how a model decides that two spellings of your name are the same company.
Review platforms AI search trusts
G2, Capterra and Trustpilot supply the structured opinion an assistant needs to justify naming you over someone else.
YouTube and trade press in AI Overviews
Demos, teardowns and reporting by people who do not work for you. Coverage in an outlet your buyers read outperforms your own blog.
Your own pages, where AI SEO starts
Still necessary, still not sufficient. Your site establishes the facts; everything above decides whether the model believes them.
Ordering reflects what we see in commercial categories and will differ in yours. The audit reports your real citation mix rather than assuming this one.
Hire an AI SEO agency: engagement tiers
Scope depends on how many pages you have and how contested your category is, so quotes are written after the audit rather than guessed before it.
Get a one-off AI visibility audit
Find out where you actually stand before spending anything on fixing it.
- A prompt set built around your buyers
- Answers recorded across six assistants
- Competitor mentions counted
- Every cited source listed
- A ranked fix list you can hand to any team
AI SEO services: audit, then build
The audit, plus the work to close the gaps it finds, done rather than recommended.
- Everything in Baseline
- Entity and schema build, shipped
- Comparison and alternatives pages written
- Crawler access, robots.txt and llms.txt corrected
- Re-measured against the original baseline
Ongoing AI visibility tracking service
For categories where competitors are working on this too and answers keep moving.
- The same prompt set re-run monthly
- Mention rate, competitor share and sentiment
- Assistant referral traffic reported separately
- Consensus work continued off-site
- Alerts when an answer turns against you
What Google, OpenAI and Perplexity say about AI citations
Six facts worth knowing, each taken from the vendor's own documentation and linked so you can check it rather than take our word for it.
OpenAI on AI search citations
OpenAI documents three separate agents: GPTBot, OAI-SearchBot and ChatGPT-User. They serve different purposes, so blocking one does not block the others, and a site can allow search retrieval while refusing training.
OpenAI crawler documentationGoogle on AI Overviews visibility
Google-Extended is a robots.txt token that controls Gemini and grounding use of your content. Google states it has no effect on Search ranking, which makes disallowing it a quiet way to opt out of AI Overviews.
Google crawler overviewAnthropic on AI search crawler access
Anthropic publishes its crawler identities and honours robots.txt directives per agent, so access decisions here are made in the same file as everything else rather than through a special process.
Anthropic support documentationPerplexity SEO and live answer fetches
Perplexity separates PerplexityBot, which builds its index, from Perplexity-User, which fetches a page because someone asked about it right now. The second one is the reason blanket blocks cost you live answers.
Perplexity crawler documentationMicrosoft on Copilot AI search visibility
Copilot answers are built on the Bing index, so bingbot access remains the dependency there. A site that neglected Bing for years is usually invisible to Copilot for that reason alone.
Bing Webmaster helpllms.txt and LLM SEO: does anything read it?
llms.txt is a community proposal, not a standard any major assistant has confirmed it reads. It costs an hour to publish and may help later, so treat it as cheap insurance rather than a fix.
The llms.txt proposal
Questions to ask before you hire an AI SEO agency
Usually because more sources outside your own website agree that they exist and are credible. Assistants lean on independent roundups, forums and reviews, so a brand that only talks about itself gets skipped.
Request your AI visibility audit
Tell us your category and the competitor you keep losing to. We run the prompts across all seven assistants, write down every name that comes back and every source behind it, and reply with what we found and what it would take to change it.
AI visibility audit requested
We will run your prompt set across all 7 assistants, write down every name that comes back and every source behind it, and reply with what we found. Usually within one business day, from a real person.
Nothing else happens in the meantime. If you want to add anything, reply to that email or write to hello@getrecommendedbyai.us.com.