Why your business isn’t showing up in ChatGPT, Gemini or AI search, and how to fix it
Your brand is missing from ChatGPT and other AI answers because the models, and the search indexes behind them, cannot find enough clear, consistent, third-party evidence of who you are, what you do and who you serve. The five usual causes are unclear entity information, too few mentions on sources the models trust, missing structured data, AI crawlers blocked or content hidden in JavaScript, and stale pages. Each one can be tested in an afternoon and fixed in weeks, and the same work underpins our generative engine optimisation service.
Why isn’t my brand showing up in ChatGPT or AI search results?
Your brand is not showing up because the assistant could not retrieve, or did not trust, enough evidence about you at the moment someone asked. That is a different problem from ranking, and it is why businesses on page one of Google can still be absent from an AI recommendation.
An AI answer is assembled in two layers. The first is what the model learned in training, which is fixed at a point in time and skewed towards brands that were already widely written about. The second is live retrieval: when you ask a current question, the assistant runs searches, reads a handful of pages and writes its answer from those, citing some of them.
As at 16/09/2026, each product does that retrieval differently. ChatGPT with search enabled pulls live web results through its own crawler and third-party search providers. Gemini and Google AI Mode draw on Google’s search index. Perplexity runs its own crawler and index and cites sources by default. The details change often, so treat this as the general shape rather than a specification.
The distinction that matters is between being known to the model and being cited in the answer. A model can know your name and still not cite you, because citation depends on what the retrieval step found for that specific prompt. Retrieval is where most businesses fail, and it is what you can influence in weeks rather than years.
Run the test column for each row below and you will know which causes apply before you spend anything on fixes.
| Symptom | Likely cause | The test | The fix |
|---|---|---|---|
| The assistant describes you wrongly, or confuses you with a similarly named business | Unclear entity information | Ask “What is [brand] and where is it based?” in each assistant and compare the answer with your About page | State name, category, location and offer identically everywhere, and add Organization schema with sameAs links |
| Competitors are recommended for your category and you never appear | Too few third-party mentions | Search your brand name with your own domain excluded and count the independent pages that describe you | Earn listings on industry lists, local roundups, review platforms and forums |
| You are mentioned but the details, prices or services are wrong or missing | Missing or thin structured data | Run your key pages through Google’s Rich Results Test and the Schema Markup Validator | Add Organization, LocalBusiness, Service, Product and FAQPage schema that matches the visible text |
| You are absent from ChatGPT or Perplexity but present in Google AI Mode, or the reverse | AI crawlers blocked or content unreadable | Check robots.txt, CDN bot rules and server logs for the AI user agents, then view a page with JavaScript disabled | Allow the crawlers you want and serve the main content as HTML |
| You appear for old services, old prices or an old address | Stale or undated content | Ask “What does [brand] currently offer and what does it cost?” and check the dates on the pages cited | Add visible dates, refresh the facts and republish key pages on a schedule |
Cause 1: entity clarity
Entity clarity means an assistant can state in one sentence what your business is, where it is and what it sells, and every source it checks agrees. If your website says “digital agency”, LinkedIn says “software consultancy” and a directory says “IT services”, the model has three weak signals instead of one strong one.
Start with the About page. Write the plain version: legal and trading name, category, suburb and city, what you do, who you do it for and since when. Repeat that wording, unchanged, in your footer, Google Business Profile, LinkedIn company page and every directory record you control.
Then add Organization schema to the home page with a sameAs array that points to those profiles. That links your site to the mentions of you elsewhere in a form machines can read.
Cause 2: too few third-party mentions
AI recommendations are drawn from lists, directories, review platforms, news and forum threads, so if your name only appears on your own website, there is nothing for the assistant to cite. Ask ChatGPT for the best accountants in Fremantle and it is summarising pages that already list a set of accountants, not weighing every firm’s website.
What counts is any independent page that names you in the right category. Industry association member lists, “best of” roundups from local publishers, Google and industry review profiles, Reddit and specialist forum threads, supplier and partner directories, and local press coverage all qualify. A single well-known list can matter more than a dozen links from sites nobody consults.
If a search for your brand name with your own domain excluded returns only your social profiles, this cause applies to you. It takes the longest to fix, so start the outreach early.
Cause 3: missing or thin structured data
Structured data does not get you recommended on its own, but it removes ambiguity about the facts once you are retrieved. Organization and LocalBusiness schema confirm who and where you are, Service and Product schema describe what you sell, and FAQPage schema pairs questions with answers in a form that is easy to lift into a response.
Validate it with Google’s Rich Results Test and the Schema Markup Validator at schema.org. The common failures are schema that contradicts the visible page, a LocalBusiness block with no address or opening hours, and FAQPage markup wrapped around marketing copy rather than real questions.
Cause 4: AI crawlers blocked or content unreadable
If an AI crawler cannot fetch your pages, nothing else in this article will help, and blocking is more common than most owners realise. Check robots.txt for rules on GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, PerplexityBot, Bingbot and Googlebot, and decide deliberately which you want to allow.
Then look past robots.txt. CDN, firewall and security plugin bot rules can challenge or block these user agents silently, so the crawler sees a challenge page while you see a normal site. Server logs or the CDN’s bot analytics show whether they are getting a 200 response.
Two points cause confusion. Google-Extended is a training opt-out control, not a search control: AI Overviews and AI Mode use the normal Googlebot index, so blocking Google-Extended does not remove you from Google’s AI answers. And llms.txt, a proposed file that summarises your site for language models, is optional and unproven, because as at 16/09/2026 no major AI platform has confirmed that it reads the file.
Finally, check rendering. Content that only appears after JavaScript runs may never be seen by a crawler that does not execute scripts, so view your key pages with JavaScript disabled and confirm the copy is still there. This is standard technical SEO work, and it is the fastest cause to fix.
Cause 5: stale or undated content
Retrieval systems and models both favour recent, dated sources, so a service page last touched in 2022 with no visible date loses to a competitor’s page updated last month. If an assistant quotes an old price or a service you no longer offer, this is the cause.
Put a visible published and updated date on every page that matters. Refresh the facts an assistant is likely to quote: services, locations, pricing model, team, certifications and opening hours. Republish the ten pages that carry your commercial intent on a fixed schedule, and update the dateModified field in your schema when you do, not before.
Why is my B2B SaaS company invisible in ChatGPT search results?
B2B software brands usually fail on cause 2, because the model answers “best [category] software for [use case]” from comparison lists and software review platforms, and if you are absent from those you are absent from the answer.
Settle on one category name and use it everywhere. If your site says “workforce planning platform”, your G2 listing says “HR analytics” and your founder’s LinkedIn says “people operations software”, the model cannot connect you to any of the three prompts.
Get listed and reviewed on the software review platforms your buyers already use, because those pages are cited constantly for software recommendations. Then publish the pages buyers ask for and assistants quote: honest comparison and alternatives pages, public pricing, an integrations list and case studies with a named client and a stated outcome. Software brands that hide pricing behind a demo form hand the recommendation to the competitor that publishes it.
Why aren’t my competitors showing up when I ask ChatGPT for recommendations in my industry?
Your competitors are not showing up because the answer is assembled from whatever the model retrieved for that prompt, and if the whole category lacks list and directory coverage, the assistant falls back to national names or generic advice. A recommendation for “best commercial fit-out companies in Perth” needs pages that already list commercial fit-out companies in Perth, and in many trades those pages barely exist.
That gap is the opening. The first business in a category with clear entity data, correct schema and a few credible mentions tends to become the one cited, because it is the only candidate the retrieval step can verify.
Add three competitors to your prompt log now, so you see the moment one of them starts appearing.
Why can’t people find my business when they ask AI for local services in Australia?
Local AI recommendations lean on Google Business Profile and Maps data, review counts and local directories, so a thin or unverified profile keeps you out regardless of how good your website is. When someone asks for a plumber in Sydney, an electrician in Adelaide or a physio in Perth, the assistant is largely reading the same local signals that feed the Google map pack.
Verify the profile, fill every field, set the primary category correctly, and keep hours, phone number and address identical to your website and directory listings. Ask for reviews consistently, and reply to them, because review volume and recency are among the few local signals an assistant can read directly.
Then make sure your website says which suburbs and cities you serve in plain text, with a service area page where it makes sense. Our Google Business Profile optimisation service covers the full checklist, and the same steps apply whether the customer is asking in Sydney, Adelaide or a regional town.
How can I check if my small business shows up in AI search results like ChatGPT?
Run the same 15 to 20 prompts across ChatGPT with search enabled, Google Gemini, Perplexity and Google AI Mode, record whether you are named, cited with a link, or absent, and repeat monthly. One casual chat tells you nothing, because answers vary from run to run.
Design the prompt set to cover five types. Category prompts (“best [category] in Australia”), category plus city prompts (“recommend a [category] in Perth”), comparison prompts (“[brand] vs [competitor]”), problem-led prompts (“who can help me with [problem]”) and reputation prompts (“is [brand] any good”). Write them the way a customer would type them, not the way you would search.
Run the prompts in fresh sessions with no chat history, using temporary chat or a logged-out browser where the product allows it, because personalisation skews results. Run each prompt three times per platform and record every run, not the best one.
Score each run 0 for absent, 1 for mentioned by name, and 2 for mentioned and cited with a link to your site. Capture the cited URLs whether they are yours or not, because those URLs are your target list for cause 2. Add three columns for the competitors you named earlier and score them the same way.
The spreadsheet needs one row per run with these columns: date, platform, prompt type, prompt, run number, your score, cited URLs, three competitor scores and notes. A summary tab then gives you the share of runs scoring 1 or above, the share scoring 2, and both figures per platform and per prompt type. Twenty prompts, four platforms and three runs is 240 rows a month, or two to three hours of work.
How to make your business show up in AI search engines: the fixes in order
Fix the causes in the order that pays back fastest: unblock crawlers and fix rendering in days, entity information and schema in a week, refreshed and restructured key pages with answer-first paragraphs and FAQs over a few weeks, and third-party mentions over months. Doing them in this order means the mentions you earn later land on a site that assistants can already read and understand.
This is the sequence our AI Visibility Accelerator follows as a fixed-scope sprint, and it is the sequence to follow if you do the work yourself.
What to fix on your site first
Start with the answer-first pattern on every page that targets a question: a direct answer in the first two or three sentences, then the detail, then a short FAQ block with the questions customers ask and 40 to 80 word answers. Our guide to answer-first formatting for featured snippets describes the same structure, and it works for AI answers for the same reason.
Then rewrite the About page and each service page so they state facts an assistant can quote: what the service is, who it is for, where you deliver it, what drives the cost and how long it takes. Add visible dates, add the schema described under cause 3, and confirm the crawlers described under cause 4 can fetch everything.
What to fix off your site
Off-site work is the slow part and the part that moves recommendations most. Work through the cited URLs from your prompt log and get listed on every one that accepts listings. Claim and complete industry review profiles, join the association directories in your sector, pitch local publishers whose roundups already rank, and answer questions in the forums your customers read.
Keep every profile consistent with the entity wording from cause 1. Then earn some local press, because one news article that names your category and city can be cited for years.
How to track brand mentions in AI chatbots over time
Combine the monthly manual prompt log with one monitoring tool and a GA4 view of referral traffic from chatgpt.com, perplexity.ai, gemini.google.com and copilot.microsoft.com. Each source covers a gap in the others.
The prompt log is your ground truth. Monitoring tools come in three categories: AI visibility trackers that run prompt sets at scale, brand monitoring platforms that have added AI answer coverage, and rank tracking suites with an AI module. Pick one on how well its prompts match yours, and keep the manual log running as a check.
In GA4, filter a report or exploration to session source containing those AI domains, plus claude.ai and any others in your referral list. Some AI referrals arrive without a referrer and land in direct, so the GA4 figure is a floor. As at 16/09/2026, Search Console also has a generative AI performance report that shows impressions in AI Overviews and AI Mode, but it reports impressions rather than clicks or queries, so treat it as a visibility signal rather than a traffic figure.
Report three numbers each month: the share of prompts that mention you, the share that cite your site, and AI referral sessions. Split the first two by platform, because the platforms move at different speeds.
What are the best practices for improving on-site SEO and generative engine optimisation together?
The foundations are shared, so do them once: crawlable pages, a clear entity, structured data and a fast mobile load serve Google rankings and AI answers alike. Then add the GEO layer, which is quotable, dated, answer-first content and off-site mentions.
In practice that means:
- One category and one description of the business, used everywhere.
- Crawl access for the search and AI crawlers you want, checked in logs, not just robots.txt.
- Schema that matches the visible text on every commercial page.
- A direct answer in the first paragraph of every page that targets a question.
- Visible dates and a refresh schedule for the pages that matter.
- A standing list of the pages that assistants cite in your category, and a plan to appear on them.
When to hand AI visibility to an agency
Hand it to an agency when the prompt test shows you absent across every platform while competitors appear, when you cannot find the reason a crawler is being blocked, or when the off-site work needs outreach you do not have time to run. The diagnosis is simple enough to do yourself and causes 4 and 5 are quick fixes, but causes 1 to 3 need technical and content work together, and cause 2 needs months of effort.
Judge any agency on evidence: a before-and-after prompt test for a client, which of the five causes it thinks applies to you and why, who does the work and where, and what happens at the end of the engagement.
PWD, established in 2007, does this work in-house from West Leederville, Perth, with ISO 9001 and ISO 27001 certification and Google Premier Partner status. AI brand monitoring is part of the GEO service, so the three numbers above are reported to you.
Frequently asked questions
How do I get my business on ChatGPT recommendations in Australia?
Make your business easy to verify and easy to cite. State your name, category, location and offer identically on your site, Google Business Profile and directories, add Organization and LocalBusiness schema, allow the AI crawlers in robots.txt and your CDN, put visible dates and answer-first paragraphs on your key pages, and earn mentions on the lists, review platforms and forums that ChatGPT already cites for your category.
Can I pay to appear in ChatGPT answers?
You can buy ads in ChatGPT, but you cannot buy a place in the answer itself. As at 16/09/2026, ChatGPT ads are available to Australian advertisers, appear as labelled sponsored placements separate from the response, and are shown only to users on the free and low-cost Go plans, not on the paid Plus, Pro, Business or Enterprise plans. The sources an assistant cites are chosen by retrieval, so the five fixes in this article are the only route into the answer.
Does ranking first on Google mean I will appear in AI answers?
No, because ranking and citation use different logic. Google ranks your page for a query, while an AI assistant retrieves a handful of pages for a prompt and then decides which facts and sources to include based on clarity, consistency and corroboration elsewhere. A page-one ranking helps retrieval find you, but without entity clarity, schema and third-party mentions the assistant may read your page and cite a competitor.
How long before AI answers change after fixes?
Weeks for retrieval-based answers and longer for model knowledge. Once crawlers can read a refreshed page and the schema is live, ChatGPT search, Perplexity and Google AI Mode can pick up the change within a few weeks of a recrawl. Answers drawn from training data change only when a model is retrained, which is why our generative engine optimisation service treats the entity and mention work as a standing programme rather than a one-off.
Find out why AI assistants are not recommending you
The AI Visibility Accelerator is a fixed-scope sprint delivered in-house in Perth over 14 business days. It starts with a baseline of where you appear in AI answers now, works through the on-site fixes above, including schema, page optimisation and internal linking, then re-checks the result after 30 days and hands over to monthly monitoring.
If you would rather see the full service first, our generative engine optimisation service page explains how GEO and SEO run together on one plan. Find out why AI assistants are not recommending you: start with the AI Visibility Accelerator, or call 08 6146 0195 to talk it through.
