Your business does not show up in ChatGPT or Perplexity because the assistants do not know it exists, or because they cannot find a page that answers the buyer's question with your name on it. There is no form to fix that. Assistants build answers from search indexes, from the passages on your pages and from what review sites, directories and other companies say about you. This guide by Scalehackerlab, a Growth-as-a-Service company in Miami that sells this work as SEO & AI search, covers how the assistants pick businesses, how to test your own, what to change, how long it takes, and the baseline we measured for ourselves on October 28, 2026. The deeper guide is our GEM guide.
Do AI assistants have a submission form?
No. There is no place to submit your business to ChatGPT, Claude or Perplexity, and nobody can sell you a listing. When a user asks for a recommendation, the assistant runs web searches, reads the pages it finds and names the businesses those pages mention. Google says the same about its own AI features: there are "no additional requirements to appear in AI Overviews or AI Mode" beyond being indexed and eligible for a snippet, per its AI features documentation. So your visibility in AI answers is decided on pages you control (your website) and pages you influence (review sites, directories, lists, press). The only "form" is indirect: let the crawlers read your site, and get your name, category and city onto the sites the assistants trust.
How do ChatGPT, Claude and Perplexity choose which businesses to name?
They retrieve pages from a search index, pull out the passages that answer the question, and name the businesses in those passages, preferring the ones mentioned on several sites. Three facts shape this. First, being talked about matters more than links: across 75,000 brands, Ahrefs (2025) found branded web mentions correlate 0.664 with appearing in Google's AI Overviews, against 0.218 for backlinks. Second, you do not need to rank first: in Semrush's study (2025), ChatGPT search cited pages ranking in positions 21 and beyond about 90% of the time, and Ahrefs (2026) found only 37.9% of URLs cited in AI Overviews also sat in the top 10. Third, assistants are picky with local businesses: ChatGPT recommended just 1.2% of more than 350,000 brand locations, against a 35.9% appearance rate in Google's local 3-pack, according to SOCi (2026). Claude with web search works the same way: it searches, reads and summarizes, so the same rules apply.
How do I test whether my business shows up?
Ask the assistants the questions your buyers ask, several times each, and count three things: mentions (your name appears), citations (your site is linked) and recommendations (you are named as a choice). Ten prompt templates for a local or small business; replace the brackets.
- "What should I look for when hiring a [category] in [city]?"
- "Who are the best [category] providers in [city]?"
- "How much does [service] cost for a small business in [state]?"
- "Recommend a [category] for a [industry] company with [number] employees."
- "Compare the top [category] options in [city] for a small business."
- "Is [your business name] a good [category]? What do reviews say?"
- "I need a [category] that works in English and Spanish in [city]. Who should I call?"
- "What questions should I ask a [category] before signing a contract?"
- "[Category] in [city] for under $[budget] a month: what are my options?"
- "Which [category] near [neighborhood] can start this month?"
Run each prompt three or four times on each assistant, with web search on, because answers change from run to run. For every answer, record who was named, linked and recommended, then divide by the total number of answers. That gives a mention rate, a citation rate and a recommendation rate you can repeat next month.
What should my website say so an assistant can quote it?
Say who you are, what you do and where in the first paragraph of every important page, then answer the buyer's questions in plain, visible text. Assistants quote passages, so each passage has to stand on its own. Seven rules:
- Name the business and its category in the opening paragraph: "Scalehackerlab is a Growth-as-a-Service company in Miami that..."
- Use question headings that match what buyers ask, with a complete answer directly below each one.
- Use lists and tables for prices, steps and comparisons. Microsoft's Bing team recommends headings, Q&A formats, lists and tables, and says: "Don't hide important answers in tabs or expandable menus" (Microsoft, 2025).
- Show a visible "Updated on" date and keep it honest. AI assistants cite content that is 25.7% fresher than organic results, per Ahrefs (2025).
- State prices as text, with the currency and what is included.
- Publish your FAQ as visible questions and answers, outside accordions and images.
- Allow the answer-engine crawlers. OpenAI's bots documentation explains that OAI-SearchBot surfaces sites in ChatGPT search; blocking GPTBot only opts out of training. Check your firewall and CDN as well as robots.txt.
Which outside sites do assistants trust?
The sites that already rank for "best [category] in [city]": review platforms, directories and published roundups. In our own test the assistants relied on agency roundups, Clutch profiles and lists that agencies published about themselves. For a local business the same role is played by Google Business Profile, Yelp, industry directories and trade association member lists. Two rules follow. Mentions matter more than links, as the Ahrefs correlation shows, so a roundup that names you without linking still counts. And your profile data has to match your website: same name, category, city and prices, or the assistant may not connect a "digital agency" on a review site with the "Growth-as-a-Service company" on your homepage. Earn reviews where your category collects them (ChatGPT-recommended locations averaged 4.3 stars in the SOCi study), claim the profiles, and ask the publications that list your competitors to include you.
How long until it changes?
Weeks to months, and it moves in steps. On-page changes can be picked up as soon as the page is re-crawled, usually days to a few weeks for a small site. Third-party mentions take longer: a profile update may take a month to appear, and a roundup pitch may take a quarter to land. Citation patterns are also volatile; Semrush (2025) tracked Reddit falling from roughly 60% of ChatGPT responses to about 10% in six weeks. That is why the test above should be repeated monthly, with the same prompts, so you measure a trend rather than one day's answer. Anyone who promises a guaranteed position in ChatGPT is selling something that does not exist.
What we measured on our own site
On October 28, 2026 we ran 14 buyer prompts 4 times each on Claude with web search and ChatGPT with web search, 56 answers in total. The prompts covered what to look for when hiring, cost, the best providers in Miami, comparisons, bilingual providers and budget. Scalehackerlab was mentioned in 7.1% of answers, cited (linked) in 7.1% and recommended in 5.4%. On the 8 comparison answers (agency vs in-house vs Growth-as-a-Service, AI agents with senior review) we appeared 0 times, even where the question described our offer word for word. On the Miami prompts we appeared in half the answers, and ChatGPT ranked us first twice, citing our locations page. The pages the assistants relied on were agency roundups, Clutch and self-published agency lists. And our product name "Scale AI-hub" was read as Scale AI, the data-labeling company, which has no relation to us.
The fixes we made in the days that followed:
- The brand is named in the opening paragraph of every page and post.
- The FAQ became question headings with the answer visible below each one.
- Price tables are now real HTML tables.
- Every page shows a visible "Updated on" date.
- The comparison page was expanded.
- This guide was published.
| Metric | Baseline, Oct 28, 2026 | 90-day target (a goal, to be re-measured in late January 2027) |
|---|---|---|
| Mention rate | 7.1% | 20% |
| Citation rate | 7.1% | 20% |
| Recommendation rate | 5.4% | 12% |
| Comparison prompts | 0% | 15% |
| Local Miami prompts | 50% | 70% |
The second column is a target. We will repeat the same test, same prompts and same assistants, and republish the numbers here after 90 days, whether they moved or not.
What does an AI Search Sprint include?
The AI Search Sprint is a $499 add-on to any Scalehackerlab plan that does for your business what we did for ours: a baseline of 14 to 20 buyer prompts across the assistants, with mention, citation and recommendation rates recorded; the on-page fixes above applied to the pages that matter; profile alignment on the third-party sites your category uses, so that name, category, city and prices match everywhere; and a re-measurement with the same prompts at the end. Our AI agents do the research and drafting, and a senior marketer reviews every page before it ships. Plans start at $199 a month with no setup fee, month to month; see pricing. To know where you stand first, the Free Growth Assessment returns a 3-page report within 48 hours, with no credit card.