There's No Submit Button for ChatGPT. What Actually Gets You Cited.

Scammers are selling a product that cannot exist. The pitch arrives by phone, email, and cold outreach: for $499 we'll submit your business to ChatGPT, Perplexity, and other AI platforms so you appear in AI search results. Some versions of the pitch claim the seller is an official OpenAI partner who can get the business into ChatGPT's database.

In June 2026, the US Small Business Administration flagged this pattern at its Tech Talk events. Futurism documented the pitch templates and customer losses separately. Our July 4, 2026 research update in `smb-ai-visibility-demand-signals-2026-05.md` (session 67) synthesized these sources with practitioner reports from surmado.com: roughly 1 in 4 small businesses have been targeted by some variant of this service. Fees range from $99 to $499.

The pitch works because the underlying problem is real. Business owners know their competitors appear in AI recommendations and they don't. That's a genuine gap with genuine consequences. The scam extracts money from the urgency without touching the actual problem.

Why there is no submission form

OpenAI, Anthropic, Google, and Perplexity have no mechanism for business owners to submit their listing for AI inclusion. No paid placement system. No partner program for business database entries. No intake form.

This follows from how these AI systems work.

ChatGPT builds recommendations from two sources. The first is training data: what appeared across the public web when the model was trained. Wikipedia entries, directory profiles, review platform pages, news coverage, and website content get incorporated into the model's knowledge. No third-party payment changes what was indexed on the internet before the training cutoff.

The second is live web retrieval. When ChatGPT searches the web to answer a recommendation query, it retrieves pages that organic search surfaces for that query. A business appearing on Yelp, BBB, and category comparison sites that rank in Google organic search lands in the retrieval set. A business that paid $499 to a third party and did nothing else does not.

There is no third channel. A third party that collected your $499 and "submitted" your business either created a low-quality directory listing or did nothing at all -- and has no access to the mechanisms that actually determine AI recommendations.

Why so many businesses are targeted

ChatGPT recommends only 1.2% of local business locations on consumer-intent queries. That figure comes from SOCi's 2026 Local Visibility Index across 350,000 locations -- a large-sample study, not a practitioner estimate. Perplexity's rate is higher at 7.4%, but that still means more than 9 in 10 businesses are absent from AI recommendations at any given moment.

Business owners experience this directly. They search for their category in their city and see competitors named by AI. They know the problem is real. The $499 pitch arrives as a solution to something they have already confirmed is happening.

That's the vulnerability the scam targets. Our session 67 research noted that the "official OpenAI partner" framing appears specifically in cold outreach targeting contractors, restaurants, and professional services -- categories where owners are active on AI platforms and have already noticed the citation gap. OpenAI has no such partner program for business listings.

What the research shows actually moves the needle

Our research identifies two pre-flight checks that diagnose the most common reasons local businesses are absent from AI recommendations. Neither involves payment to a third party.

**Review volume alongside star rating.**

In our July 6, 2026 methodology update (`2026-07-06-review-volume-threshold.md`, session 69), we formalized a gap in how most fix plans handle reviews: they check star rating quality -- businesses below 4.0 stars are excluded from ChatGPT's recommendation candidate pool entirely -- but do not check review volume as a second, independent threshold.

The Seer Interactive and Trustpilot study published May 2026 analyzed 804,491 AI responses across ChatGPT, Gemini, Perplexity, and Google AI Mode -- 1,926 brands, 15,000+ distinct prompts. Citation rates by review tier:

| Review profile | AI citation rate | |---|---| | No review presence | 1% | | Minimal presence, low volume | Intermediate | | Active profile, moderate reviews | 53.5% | | 80+ reviews, actively managed | 75.3% |

The study measures Trustpilot presence specifically -- the mechanism captures review platform visibility combined with volume and response cadence, not Google reviews in isolation. But the directional finding is clear: a business at 4.7 stars with 9 reviews is near-invisible, not because its quality signals are wrong but because its volume signal is too weak. Infrastructure fixes -- schema, directory listings, NAP cleanup -- return significantly less when review volume is this low.

The practical threshold from the July 6 rec: under 20 reviews, volume is a prerequisite fix. Under 80 reviews, infrastructure work and review building should run in parallel. The 80-review mark is where the Seer data shows citation rate reaching the 75%+ tier.

**Third-party platform presence over brand-site optimization.**

On May 23, 2026, OpenAI made GPT-5.5 Instant the default ChatGPT model for free users. Our June 16, 2026 analysis (`2026-06-16-gpt55-chatgpt-citation-shift.md`, session 49) tracked the mechanism through Writesonic's study of 11,469 web search results and 1,257 citations before and after the switch.

GPT-5.4 used Google's `site:` operator to proactively look up specific brand domains in 40.5% of its web search queries. GPT-5.5 uses `site:` in only 12.6% of queries. Without that explicit brand-seeking behavior, the candidate pool defaults to whatever organic search returns: review platforms, comparison sites, journalism, forum threads. SISTRIX analyzed 3.8 million German-language ChatGPT responses after the switch and found 47% of citations had redistributed -- Reddit alone gained +7,007 citations per 10,000 responses, a 59% increase.

For a local service business, this means third-party presence is now the primary lever for ChatGPT citation. Your Yelp listing, BBB profile, Angi presence, HomeAdvisor page. Not because these contain better content than your website -- because these are the sources ChatGPT now retrieves when it does not proactively seek out your brand directly.

NAP consistency is the structural prerequisite for all of it: business name, address, and phone number matching exactly across all directory listings. Inconsistent NAP -- two different phone numbers, three variations of the business name -- fragments the entity signal across every platform simultaneously.

The actual fix sequence

A business absent from ChatGPT recommendations for its category and city has a diagnosable problem. The diagnosis typically surfaces one or more of these:

1. Star rating below 4.0 -- ChatGPT exclusion, binary, not a ranking penalty 2. Review count below 20 -- volume signal too weak to compete even with correct infrastructure 3. Missing or incomplete directory presence on Yelp, BBB, and industry-specific platforms 4. NAP inconsistencies across listings that fragment the entity signal 5. AI crawler blocks in robots.txt -- GPTBot or PerplexityBot blocked, preventing retrieval entirely

None of these require payment to a third party. Most are free to fix. The review volume gap is addressed by requesting reviews from satisfied customers systematically. The directory gap is addressed by claiming and completing existing profiles. NAP problems are fixed by editing listings to match.

The $499 submission service cannot address any of these gaps because it has no access to the mechanisms creating them.

Signal Check at sourcepull.ca runs live cross-platform queries across ChatGPT, Perplexity, Gemini, and Claude and identifies which of these specific gaps apply to your business. That's a different kind of answer than a submission service that collects a fee and leaves the underlying problem exactly as it was.