DEV Community: tomy romy

AI Can Write Like a Human — But Can It Find YOUR Product? The Invisible SaaS Problem

tomy romy — Sat, 27 Jun 2026 19:04:05 +0000

Everyone is talking about AI generating content, replacing jobs, and writing code. But there's a problem nobody is discussing:

AI search engines are now WHERE people discover software. Over 40% of product discovery starts in ChatGPT, Perplexity, or Claude — not Google.

But here's the twist: the same AI that can write a perfect essay CANNOT find most SaaS products. Why?

Because LLMs don't crawl the web like Google. They resolve entities from structured data. If your product doesn't have:

JSON-LD structured metadata
An llms.txt file
AI-friendly robots.txt
A clear machine-readable identity

...then you're invisible. Not ranked low. INVISIBLE.

I tested 10 well-known SaaS tools by asking ChatGPT and Perplexity to recommend them. 8 out of 10 were completely missing from AI responses. These tools had perfect Google rankings.

This is called Generative Engine Optimization (GEO) — and less than 1% of software products are doing it.

I built CitableHub to solve this. It's a free platform that creates AI-optimized profiles so your product becomes citable by LLMs.

The irony? AI is powerful enough to replace search — but most products aren't structured enough for AI to find them.

Try it: ask ChatGPT right now "What's the best [your category] tool?" — did your product show up?

https://citablehub.com
https://citablehub.com/scan

From SaaS to AaaS: The Ultimate Guide to Surviving the Era of AI Agents

tomy romy — Sun, 21 Jun 2026 16:03:43 +0000

Over the last two decades, the dominant software model has been SaaS (Software as a Service), primarily designed by humans to be used by humans through visual interfaces
. However, in 2026, we are witnessing the most profound shift in information infrastructure since the invention of hyperlinks: the transition to AaaS (Agent as a Service)
.
In this new paradigm, software is no longer just a visual destination; it is a "capability" that AI agents—such as ChatGPT, Claude, or Perplexity—consume autonomously to execute tasks on behalf of users
. If your software is not readable by these machines, you essentially do not exist for the 40% of buyers who now start their search in AI tools instead of Google
.
Here is how to lead this technical and strategic migration.
**
**What Exactly is AaaS?
Agent as a Service (AaaS) represents the evolution of APIs into autonomous components
. While SaaS focuses on human user retention and traditional "blue link" SEO, AaaS is built to be discovered, understood, and cited by Large Language Models (LLMs)
.
The key to AaaS is entity resolution
. For an AI agent, your product is not a URL with a pretty design; it is a set of structured data, technical capabilities, and verifiable evidence that it can process in milliseconds

Advantages of Migrating to an Agent-Based Model
Explosive Conversion: AI-referred traffic converts 4.4 times better than traditional organic traffic from Google
.
Operational Efficiency: AI agents are tireless; they run automations, monitor metrics, and trigger workflows continuously, increasing the usage of your features and APIs
.
Distributed Authority (Brand Gravity): By being cited by multiple models, you generate a network effect where the AI identifies you as the natural leader of your category
.
High-Intent Discovery: Users arriving from a ChatGPT recommendation have already been pre-qualified by the model
.
The Major Technical Challenges
Machine Invisibility: Most current SaaS platforms are JavaScript-heavy and lack structured data, making them unreadable for AI crawlers
.
Token Budgets: Agents have limited context windows
. If your documentation is messy, the agent will waste tokens and prefer to cite a more efficient competitor
.
Non-Human Identity (NHI) Security: Providing direct access to agents requires strict governance to avoid "identity drift" and ensure bots act under controlled permissions
.
Steps to Migrate from SaaS to AaaS
1. Perform a GEO (Generative Engine Optimization) Audit
Before changing any code, you need to know how "citable" you are
. Use tools to measure your Citability Score across dimensions like Identity, Evidence, Trust, Freshness, and Classification
. You must also identify if your robots.txt is accidentally blocking bots like OAI-SearchBot or PerplexityBot
.
2. Implement the llms.txt Standard
Create an /llms.txt file at the root of your domain
. This Markdown file serves as a roadmap for machines, telling agents exactly what to read and which documentation is the source of truth, saving them from processing unnecessary HTML
.
3. Structure Your Identity with JSON-LD
Do not rely on the AI to guess what your product does
. Implement SoftwareApplication and Organization schemas with complete fields, including pricing, features, founder information, and verified social links
. This allows the AI to resolve your entity without hallucinations
.
4. Create Skills Documentation (skill.md)
APIs describe endpoints, but not business logic
. For AaaS, you need files that explain to the agent when and how to use a function, providing real-world workflow examples
.
5. Connect via MCP (Model Context Protocol)
The MCP protocol is the "USB-C for AI"
. Implementing an MCP server allows tools like Claude Code or Cursor to "talk" directly to your SaaS and execute programmatic actions without you having to build custom integrations for every model

Conclusion: The Future is Cited, Not Searched
The era of ten blue links is ending
. In the AaaS ecosystem, the winner is not the one with the most backlinks, but the one who is the most reliable and easiest source of truth for the AI to cite
.
If you want to start today for free, visit [CitableHub]
. You can run a Free GEO Scan to audit your visibility and obtain your Launch Kit, which automatically generates your llms.txt file, your JSON-LD schemas, and your plugin manifest so that AI agents finally start recommending you
.

"Your SaaS Is Invisible to AI Search — Here's How to Fix it"

tomy romy — Mon, 15 Jun 2026 19:53:10 +0000

"ChatGPT, Perplexity, Claude, and Gemini are replacing Google. If your product isn't structured for AI discovery, it doesn't exist. Learn about Generative Engine Optimization (GEO) and how to make your software citable."

The Search Revolution Nobody Is Talking About
Here's a number that should concern every founder: over 40% of product discovery queries now start in AI-powered tools — ChatGPT, Perplexity, Claude, Gemini, and AI Overviews — instead of Google.

When someone asks "What's the best project management tool for remote teams?", the answer doesn't come from a list of blue links anymore. It comes from an AI that synthesizes information from structured sources, trusted references, and machine-readable data.

If your product isn't in that synthesis, you don't exist.

The Problem: Traditional SEO Is Not Enough
Traditional SEO optimizes for crawlers — programs that index keywords, backlinks, and HTML structure. But LLMs don't work like crawlers. They work like researchers:

Traditional SEO AI Discovery (GEO)
Keywords & backlinks Structured entity data
Meta descriptions Machine-readable profiles
Page rank Trust signals & evidence
SERP position Citability score
Indexing bots LLM training & RAG
This new paradigm is called Generative Engine Optimization (GEO) — the practice of making your product understandable, citable, and recommendable by AI systems.

What Makes a Product "Citable" by AI?
After analyzing hundreds of products and how LLMs reference them, we identified 5 pillars of AI citability:

🏢 Entity Identity
Does your product have a clear, structured identity? Name, category, URL, creator — all in machine-readable format (JSON-LD, structured data).
📊 Evidence & Trust
Can AI systems verify your claims? Open-source code, documentation, public APIs, demo links, privacy policies — these are trust signals for LLMs.
🎯 Classification
Is your product correctly categorized? LLMs need to know what kind of tool you are to recommend you in the right context.
🕐 Freshness
When was your product last updated? LLMs prefer actively maintained products with recent version numbers and update dates.
🔗 Structured References
Does your product exist in multiple trusted sources? Directory listings, llms.txt files, RSS feeds, and API endpoints all contribute to discoverability.

Introducing CitableHub: An AI Visibility Platform
CitableHub is a free platform that helps digital products become discoverable by AI search engines.

Here's what it does:

AI-Optimized Profiles — Creates structured, JSON-LD enriched project pages with semantic metadata that LLMs can parse and cite
Citability Score — A multi-factor score (0-100) measuring how well AI systems can discover and reference your product, broken down into Identity, Evidence, Trust, Freshness, and Classification sub-scores
Discovery AI Analyst — A built-in AI chat that queries the CitableHub index to match user needs with listed products
MCP Gateway — A Model Context Protocol server that exposes CitableHub data directly to AI agents and tools
GEO Audit — Automated analysis of your product's Generative Engine Optimization readiness
Badge System — Embeddable SVG badges showing your citability score on your own site
llms.txt & Machine-Readable Feeds — Standardized feeds specifically designed for AI crawlers
It's 100% Free
CitableHub is completely free for listing projects. No paywall, no freemium limitations. List unlimited projects with full AI discovery features.

The only paid feature is an optional one-time GQI Boost for products that want amplified visibility — but the core platform is free forever.

How It Works: A 5-Minute Setup
Step 1: Submit Your Project
Go to citablehub.com and click "List Your Product Free". Fill in your project details — or let the AI Autofill scan your website and pre-populate everything.

Step 2: Complete Your Profile
The more data you provide, the higher your Citability Score. Add trust links (documentation, GitHub, demo URL), social profiles, differentiators, and target audience.

Step 3: Get Your Citability Score
CitableHub calculates your score across 5 dimensions and gives you actionable recommendations to improve.

Step 4: Embed Your Badge (Optional)
Add a citability badge to your website:

Step 5: Monitor & Improve
Track your visibility in the dashboard. See how AI systems interact with your profile through analytics.

The Technical Architecture (For the Curious)
CitableHub is built with modern web technologies and designed for AI consumption:

Structured Data: Every project page includes SoftwareApplication, BreadcrumbList, and WebPage JSON-LD schemas
llms.txt: A machine-readable manifest at /llms.txt following the emerging standard for AI-friendly sites
RSS Feed: Full-content RSS at /feed.xml for AI crawlers that consume syndicated content
MCP Gateway: JSON-RPC endpoint at /api/mcp implementing the Model Context Protocol with 5 tools: search_projects, get_project, list_categories, get_stats, and compare_projects
OpenAPI Spec: Full API specification at /openapi.json for AI agents
AI Plugin Manifest: /.well-known/ai-plugin.json for ChatGPT and compatible AI systems
Programmatic SEO: Auto-generated /best/[category] and /compare/[a]-vs-[b] pages
Open Source: The full codebase is available on GitHub
Tech Stack
Next.js 14 · React · Tailwind CSS · Prisma ORM · PostgreSQL · Stripe · NextAuth.js

GEO Checklist: Make Your Product AI-Visible Today
Whether you use CitableHub or not, here's a checklist to improve your AI discoverability:

Add SoftwareApplication JSON-LD to your homepage
Create an llms.txt file at your domain root
Add /.well-known/ai-plugin.json if you have an API
Ensure your robots.txt allows GPTBot, ClaudeBot, PerplexityBot
List your product in AI-friendly directories (like CitableHub)
Add structured FAQ schema to your site
Include version numbers and last-updated dates
Provide a clear one-line summary of what your product does
Link to evidence: docs, GitHub, demo, privacy policy
Create comparison and category pages for programmatic SEO
The Future of Product Discovery
We're at an inflection point. The products that invest in AI visibility today will dominate discovery tomorrow. Just like early SEO adopters captured organic traffic for years, early GEO adopters will capture AI recommendations.

Your customers are already searching in ChatGPT, Perplexity, and Claude. They may never visit Google first.

The question isn't whether AI search will replace traditional search — it's whether your product will be part of the conversation.

Get Started
🔗 Website: citablehub.com
📂 GitHub: github.com/AIJAEL11/actualizacion-vip-citablehub
📖 Standard: github.com/AIJAEL11/citablehub-standard
🤖 MCP Gateway: citablehub.com/mcp
📡 LLMs.txt: citablehub.com/llms.txt
🔍 Scan Tool: citablehub.com/scan

CitableHub is built by Rafael Lopez — a founder who believes every product deserves to be discovered in the age of AI search.

SEO Didn't Die. Distribution Did.

tomy romy — Sat, 06 Jun 2026 20:05:24 +0000

SEO Didn't Die. Distribution Did.

After building multiple AI products, I kept hearing the same advice:

"Create more content."

"Publish more blogs."

"Work on your SEO."

So I did.

And something strange happened.

The content wasn't the problem.

Distribution was.

The Shift Nobody Talks About

For years, the internet worked like this:

Content → Search → Clicks → Customers

If you created useful content and optimized it well, people could eventually find you.

Today that model is changing.

Users ask ChatGPT.

They ask Perplexity.

They ask Google AI.

And increasingly, they get answers without visiting websites.

The click is disappearing.

The Real Problem

Most founders think they have a traffic problem.

Many actually have a discoverability problem.

A project can be:

Useful
Legitimate
Well designed
Solving a real problem

And still remain invisible.

Not because it lacks quality.

Because it lacks structure.

AI systems cannot recommend what they cannot understand.

What I Learned Building My Own Projects

While working on several products, I noticed something interesting.

The websites existed.

The information existed.

But the information was scattered.

No clear identity.

No structured evidence.

No defined audience.

No standardized outcomes.

No machine-readable trust signals.

To an AI system, many projects look almost identical.

A Different Question

Instead of asking:

"How do I rank higher?"

I started asking:

"How do I become understandable?"

That question changed everything.

The Idea Behind CitableHub

CitableHub started as an experiment.

What if every project could have a structured profile designed not only for humans, but also for AI systems?

A profile that clearly defines:

What the product does
Who it serves
What outcome it creates
Where it fits
When it should be recommended
When it should not

Not another directory.

A discoverability layer.

Building for AI

We began implementing:

Structured data
JSON-LD
AI crawler support
llms.txt
RSS feeds
Entity classification
Discoverability scoring
IndexNow integration

Not because users care about those things.

Because AI systems do.

What Happens Next?

I don't believe SEO is dead.

I believe discoverability is changing.

Ranking pages is no longer enough.

Being understood matters more.

The winners of the next decade may not be the loudest companies.

They may be the easiest to understand.

The Experiment

Today we are testing a simple idea:

Can structured entities outperform traditional visibility tactics in an AI-first world?

Instead of chasing attention, we are focusing on clarity.

Instead of publishing more noise, we are focusing on better signals.

Instead of asking how to rank, we are asking how to become understandable.

Final Thought

If traditional SEO was about ranking pages,

AI discoverability is about building entities.

That's the experiment I'm running right now.

If you're building something valuable and wondering why nobody can find it, maybe the problem isn't your content.

Maybe it's your discoverability.

What do you think?

Are AI systems changing how new products get discovered?

I'm documenting the experiment publicly as we continue building CitableHub.

https://citablehub.com