Your SaaS Is Invisible to ChatGPT — Here's Why

Table of Contents

When someone asks ChatGPT, Claude, or Perplexity to recommend a product in your category, does your SaaS come up? For the vast majority of products, the answer is no. Not because AI has evaluated your product and rejected it — but because AI doesn't know your product exists.

This is the difference between being judged and being invisible. Most SaaS founders assume their products are competing in AI search results. They aren't. They're not even in the conversation.

The Scan: 41 SaaS Products, One Question

To understand the scale of this problem, I scanned 41 SaaS products across multiple categories using AImpactScanner (opens in new tab), evaluating each one against the 27 factors in the MASTERY-AI Framework that determine whether AI systems can find, understand, and recommend a product.

The scan wasn't limited to obscure startups. It included established products with real user bases, active marketing, and solid traditional SEO. These are products that rank on Google, have paying customers, and would consider themselves competitive in their markets.

The question was simple: if an AI system were asked to recommend a solution in each product's category, could it find and cite these products?

The Results: 85% Invisible

Of the 41 products scanned, only 10 — roughly 24% — had the technical and content signals that AI systems need to discover and recommend them. The remaining 31 products were, for practical purposes, invisible to AI search.

The breakdown tells a clear story:

• 23 products had no llms.txt file at all — no machine-readable signal telling AI what the product does or who it's for
• 6 products had llms.txt files that were broken, malformed, or returned errors
• 2 products had llms.txt files so bloated that AI systems couldn't meaningfully parse them
• 10 products had well-structured AI signals and were likely to surface in AI recommendations

This isn't a gap in product quality. Many of the invisible products were genuinely excellent tools. The gap is in AI discoverability — and it's almost entirely a technical and content problem that can be fixed.

Why Most SaaS Products Are Invisible to AI

AI search engines don't browse the web like humans. They don't click through product pages, read testimonials, or try free trials. They rely on structured signals — technical markers, content patterns, and machine-readable context — to understand what a product does and whether to recommend it.

The most common reasons products fail this test:

No Machine-Readable Product Description

Most SaaS websites are optimised for human visitors: polished landing pages with hero images, feature grids, and call-to-action buttons. But AI systems can't interpret a hero image or a "Get Started Free" button. They need structured, text-based descriptions that explicitly state what the product does, who it's for, and what problems it solves.

Missing Structured Data

JSON-LD schemas — Product, SoftwareApplication, Organization — give AI systems a structured data layer to parse. Without them, AI has to infer product details from unstructured page content, which it often fails to do correctly or completely.

No llms.txt File

The llms.txt specification is a plain-text file at your domain root that provides AI systems with a structured overview of your product. Think of it as a robots.txt for AI — except instead of telling crawlers what not to index, it tells AI what your product is and when it's relevant.

23 of the 41 products scanned had no llms.txt file. For those products, AI systems have to piece together product information from scattered web pages — a process that frequently results in incomplete or incorrect understanding.

Weak Authority Signals

Even when AI can find your product, it needs reasons to trust it. Named authorship, consistent brand identity across platforms, third-party mentions, and verifiable credentials all contribute to the trust signals AI uses to decide whether to recommend you.

The Role of llms.txt

The llms.txt file has emerged as one of the most impactful signals for AI product visibility. It provides a single, structured document that answers the questions AI systems ask when deciding whether to recommend a product:

• What does this product do?
• Who is the target audience?
• What problems does it solve?
• How does it compare to alternatives?
• What are the key features?
• What is the pricing model?

The products in our scan that scored highest on AI visibility all had well-structured llms.txt files. Not coincidentally, these were also the products most frequently cited in AI search responses when we tested relevant queries across ChatGPT, Claude, and Perplexity.

But having a llms.txt file is not enough on its own. The 6 products with broken files and the 2 with bloated, unparseable files demonstrate that quality matters as much as presence. A poorly structured llms.txt can be worse than none at all, because it may cause AI systems to form an incorrect understanding of your product.

The Indie Hacker Problem

The scan included a separate look at 19 products from prominent indie hackers — builders who are active on Twitter/X, ship publicly, and have engaged audiences. The result was stark: zero out of 19 had llms.txt files.

This isn't because indie hackers lack technical sophistication. It's because AI search optimisation isn't yet on most builders' radar. The conversation in indie hacker communities still centres on traditional SEO, Twitter distribution, and Product Hunt launches. Meanwhile, large companies are hiring dedicated AI SEO specialists at six-figure salaries to solve this exact problem.

Traditional SEO traffic is declining an estimated 40% year-over-year as users increasingly query AI systems instead of Google for product recommendations. The shift isn't coming — it's here.

The irony is that indie hackers are better positioned to move fast on this than enterprises. A solo founder can add AI visibility signals to their product in an afternoon. An enterprise needs procurement, legal review, and a committee meeting. The speed advantage that indie hackers pride themselves on applies directly here — but only if they're aware the race has started.

What to Do About It

If your SaaS product isn't visible to AI search engines, the fix is neither complex nor expensive. It's a sequence of specific, technical steps:

1. Create a Quality llms.txt File

Your llms.txt needs to be more than a dump of your sitemap URLs. It should be a curated, structured overview of your product: what it does, who it serves, how it works, and what differentiates it. Tools like LLM.txt Mastery (opens in new tab) can generate and optimise this file for you.

2. Add Product Structured Data

Implement JSON-LD schemas for your product pages. At minimum, add SoftwareApplication or Product schema with name, description, pricing, and category. This gives AI systems structured data to reference directly.

3. Audit Your AI Visibility

Run your product through an AI visibility audit (opens in new tab) to understand exactly which of the 27 MASTERY-AI factors your product is failing on. A generic "improve your SEO" approach won't work here — AI visibility has specific requirements that differ from traditional search optimisation.

4. Build Entity Authority

AI systems weight third-party signals heavily. Get your product mentioned in relevant publications, directories, comparison sites, and community discussions. Every external mention strengthens the entity signal AI uses to decide whether your product is trustworthy enough to recommend.

5. Test AI Responses Directly

After implementing these changes, test directly. Ask ChatGPT, Claude, and Perplexity to recommend products in your category. Note whether you appear, how you're described, and what competitors are cited instead. This direct testing gives you immediate feedback on whether your signals are working.

The Window Is Open — For Now

The most important finding from this scan isn't that 85% of products are invisible. It's that the 15% who are visible face almost no competition for AI citations in their categories.

When AI systems have limited data about a category, they lean heavily on the few products that have clear, structured signals. Early movers in AI visibility don't just get recommended — they get recommended disproportionately often, because there's nobody else for AI to suggest.

This window won't stay open indefinitely. As awareness grows, more products will implement AI visibility signals, and competition for AI citations will intensify. The products that establish themselves now will have the advantage of accumulated authority and citation history.

The question isn't whether to optimise for AI search. It's whether you'll do it while the field is nearly empty, or after your competitors have already claimed the territory.