Our mission
The AI Internet needs standards.
And every business deserves a place in it.
Your hours. Your services. Your products. You should have a clear way to give those facts to AI, whatever the size of your business. That’s why we’re building X-Protocols.
The trust problem
The Greatest Problem of the AI Era.
An answer from ChatGPT or Gemini can sound completely right and still be wrong about your business. Publishing more articles does not fix that.
Look inside “Rank higher in ChatGPT in 7 days”.
Maybe you have seen that ad on Instagram or Facebook. Before you pay, look at what the service actually does for the money.
If it picks the topics, writes the articles and publishes them on your website, you are paying for articles. Good articles can help you. They are not a place in an AI answer, and no service controls what AI recommends.
If you have already paid for one, you may know how it ends. You expected something to change for your business. What arrived was a draft to publish. Then you read it and think: this has nothing to do with my business. Nobody gave the AI writer your real services, so it filled the gaps by guessing.
Illustrative example
You repair car electronics. The article they wrote for you says you also fit tyres and offer roadside assistance 24 hours a day.
It sounds professional. It describes services you don’t provide.
The visits went up. Then they disappeared.
Useful articles can bring visitors. If the visitors later stop coming, that on its own does not tell you why. It is also not proof that Google punished the articles for being written by AI.
Google says it looks at whether content is good, not at whether a person or an AI wrote it. Writing with AI can be useful when it is accurate, original and about the right subject. Google’s guidance on AI content (opens in a new tab)
Publishing pages in bulk, mainly to move up in search and with little value for the reader, can break Google’s spam rules. A site that breaks them can lose its ranking or disappear from the results. Google’s spam policy (opens in a new tab)
If somebody charges you for AI recommendations, ask them for proof. The number of articles they published is not proof.
Why we’re building X-Protocols
Give AI the facts. Ready to use.
Your real business should be the starting point for what AI says about you.
You publish what you actually do: your services, your locations, your opening hours and your products. X-Protocols keeps those facts in one place that you control, and records who published them and when they last changed.
When an AI system reads that source, your details are already sorted and labelled. It does not have to work them out from the design of your website. A service written down clearly is something it can match to a real customer question.
Our aim is to make your own information easy to find and easy to use. Does that get your business mentioned more often? That is a question we still have to test. Signing up to X-Protocols does not buy you a ranking or authority inside an AI system.
01 / The idea, in plain English
You know your business. AI should hear from you.
An everyday question
“Are they open on Saturday?”
You changed your opening hours. An old listing somewhere did not. Someone asks AI, and gets the old answer.
The right information exists. It is just not the one that reaches people.
X-Protocols gives you one place to publish your business facts, keep them up to date, and make them easy for AI systems to find and check.
Your website tells your story. X-Protocols helps you put the facts on the record.
Why there is a free plan
A place for every business. Including yours.
A local workshop should have a way to share its facts, just like a large company.
You shouldn’t need a big marketing budget or a team of developers to take part.
That’s why X-Protocols has a free plan.
We believe the next chapter of the internet should be open to people building something small, too.
Paid plans support larger needs and help fund the work. Free access gives more businesses a place to start.
This belief shapes the product we build.
A word from the founder
“I want small businesses to have a voice in what comes next.”
I believe every business should have a fair chance to be understood online. That includes the person doing everything alone, with little time and a small budget.
It’s why I built X-Protocols, and why having a free plan matters to me.
My ambition is to help build a better foundation for how AI gets business information: clearer facts, a known source, and less unnecessary work.
There’s a lot ahead of us. I want us to earn trust by showing what we build and what we learn.
Aldin Smajkan — Founder, X-Protocols
02 / For developers, SEO specialists & researchers
The same idea. A closer look.
Behind the simple idea are specific questions: who published a fact, what changed, how can a machine read it, and what does the evidence actually show?
For developers
A source machines can work with.
Businesses publish the information. The system gives it a structure and publishes signals about where it came from and when it changed. It does not ask an AI model to decide which business claim is true.
-
Publish facts
Services, locations, opening hours and products become structured records. Machines can read fields instead of extracting every fact from a page layout.
-
Identify the source
DNS verification connects a publisher to control of a domain. That is evidence about who controls the source, not proof that every claim is accurate.
-
Track changes
Versions and update times make revisions traceable. Freshness still depends on the business keeping its information current.
-
Make it readable
APIs, discovery manifests and machine-readable pages provide routes to the records. Signed payloads support integrity checks; signatures do not establish real-world truth.
The technical goal: make the source and state of a business record easier to inspect. Whether an AI retrieves, trusts or uses it is a separate step.
For SEO & search specialists
Publishing a source is the first step.
X-Protocols gives a business another way to publish clear, structured facts alongside its website.
It does not buy placement in an AI answer. We do not promise that ChatGPT, Gemini, Claude or Perplexity will mention a business.
Google says existing SEO fundamentals also apply to AI Overviews and AI Mode. No special AI file or schema is required, and meeting the requirements does not guarantee inclusion. Google’s guidance on AI features (opens in a new tab)
We separate what is published, what is accessed and what actually appears in an answer.
- Published
- The record is available. That alone does not establish indexing or selection by an AI system.
- Accessed
- A logged request shows that an endpoint was requested. It does not prove a recommendation, customer visit or lead.
- Used in an answer
- This needs separate observation of the actual answer, its date and context. A crawler log cannot stand in for that evidence.
For researchers / An early experiment
A simpler route. An early result.
We compared three routes to the same business information: raw HTML, content with markup removed, and X-Protocols. The test used an active website and the production platform.
Measured requests
142
Raw HTML compared with X-Protocols in this task.
Modelled energy reduction
≈96.5%
Estimated reduction against raw HTML under the simulation’s assumptions.
Requests and data sizes were measured. Energy was estimated. These are results from one controlled comparison, not direct electricity readings or a saving promised for every AI query.
| Route | Requests | Reported data size | Estimated energy |
|---|---|---|---|
| Raw HTML | 14 | 1,319,531 B | 10.55 Wh |
| Markup removed | 14 | 97,394 B | 0.85 Wh |
| X-Protocols | 2 | 46,156 B | 0.37 Wh |
What this suggests
In this task, direct retrieval needed fewer requests. The model estimated lower energy use than either crawling route — about 57% lower than the markup-stripped route.
That gives us a specific idea to test further: publishing facts directly may reduce the work needed to retrieve and process them.
How to read the numbers
The three energy estimates use the same model assumptions. Actual energy use depends on the infrastructure and workload.
The cleaned-content size may describe output after processing, rather than network traffic. These figures do not establish an equivalent reduction in electricity use or carbon emissions.
X Labs / The questions ahead
Build it. Test it. Show the work.
X Labs is our research direction: studying how AI finds and uses business facts, and how to measure the difference a clearer source makes.
We want to move from a promising example to results that others can inspect and repeat.
- Accuracy: does a clearer source change the answer, compared with the same question without it?
- Updates: when a business changes a fact, how long does it take for answers to reflect it?
- Recommendations: across repeated, comparable questions, does publishing through X-Protocols change how often a business is mentioned or recommended? The energy benchmark does not answer this.
- Efficiency: do the results hold across more websites, repeated runs and different workloads?
- Reproducibility: can another person run the same method and reach a comparable result?
Your business has a place here.
Start with your facts. Be part of what comes next.