MACHINE 404 / The symptom

Why your site does not appear in AI answers

There are five reasons, they are mechanical, and you can tell which one is yours in about ten minutes. None of them is that your website is ugly, and none of them is fixed by writing more blog posts.

1. The crawler never got in

Before anything can quote you, something has to read you. A robots.txt that blocks what it did not mean to block, a WAF that challenges anything without a browser fingerprint, a server that answers people and refuses machines: any of these ends the story before it starts, and none of them shows up in any dashboard you look at.

Worth 6 of 100 on the robots file alone, and it gates the other 94.

2. Every invented address answers 200

Type a page that does not exist on your own site. If what comes back is your homepage with a 200 instead of a real 404, you have a soft 404, and it poisons everything after it: a machine now believes that every URL it guessed is a real page of yours, so nothing it learned about your structure is true.

Worth 8. It is the second heaviest single check in the scan, and the most common serious fault we measure.

3. The company is not an entity, only a design

A person reads your header and knows who you are. A model needs it stated: legal name, contact point, where you operate, and at least one profile it can cross-check to confirm you are one business and not three similar ones. Most sites describe their pages and never describe the company, which is why an assistant can summarise your services and still not name you as the supplier.

Worth 10 plus 4 plus 5. Identity is the single biggest block of points in the scan after the offer.

4. The price is a picture

A number inside an image, inside a PDF, or inside a sentence is invisible to the thing comparing you with two competitors. It does not read the number as expensive or cheap. It reads it as absent, and absent loses to any number at all.

Worth 12, the heaviest check there is. If you sell services rather than products this is almost certainly the one costing you the most, because product markup comes free with a shop and nothing comes free with a service page.

5. Everything is behind JavaScript

Most AI crawlers do not run your JavaScript. What is in the HTML when it arrives is what they read, and on a site built as an application that can be a menu, a footer and an empty div. Open your own page source and read what is actually in it before the scripts run.

What this is not

It is not a ranking trick and there is no guarantee attached to it. Nobody controls which names a model says, and anybody who tells you they do is selling something. What is controllable is whether the facts exist in a form that can be lifted, and that is measurable, repeatable and either present or absent.