# Your domain is the source. Everything else is a copy. > Why your own domain is the foundation for company data, why profiles and portals only work after that — and how machines end up picking the wrong copy. Ask yourself once: where do the facts about your business stand **bindingly**? Not where they appear everywhere — but where they apply when two statements contradict each other. For most businesses the answer is uncomfortable: in many places at once, and bindingly in none. Opening hours in a map service, services on the website, prices in a PDF, news on a social platform. Four places, four versions, four truths. ## Why your own domain is the source Your own domain is the only place where your company data **belongs to you** — and therefore the only one that can serve as a binding source. Everywhere else there is a copy. Copies aren't bad; they simply aren't the source. A directory listing, a map-service profile, a review site: all of them reproduce what they picked up at some point. When something changes, it doesn't change there — unless someone maintains it, and that is exactly what tends to slip in day-to-day work. At btlabs Core the domain is therefore not a shopfront but the **foundation**: the place where facts are maintained once and from which every other channel is served. For a small business that isn't a technical detail but the difference between maintaining once and forgetting four times. ## What happens without a clear source Without a binding source, the machine decides for itself which copy to believe — and it often decides wrong. How often was measured by the Tow Center for Digital Journalism at Columbia University in March 2025: eight AI search tools, 1,600 queries, **more than 60 per cent incorrect source attributions**. The spread between systems was wide — from 37 per cent in one to 94 per cent in another. For you that means: if your details differ slightly across four places, that is not a cosmetic problem. It is an invitation to every machine to pick the wrong variant — and then repeat it until someone corrects it. ## How a machine recognises the source It recognises it by three things, and all three sit on your domain. **First: unambiguity.** A value appears once, not three times slightly differently. Contradictions within the same website are the strongest signal that nothing here is binding. **Second: annotation.** An opening time annotated as a time value gets read. The same opening time in running text has to be interpreted. Interpreting produces errors. **Third: currency.** Something that never changes looks, to a machine, like something nobody cares about any more. A maintained place beats a complete one standing still. ## And then the rest follows — in this order Once the foundation is in place, everything else works. Before that, it only multiplies the confusion. That is the point of the order. A listing pointing at a clean source strengthens it. The same listing without a clean source strengthens a contradiction. Starting with the copies means working against yourself. Concretely, for a small business: 1. **The domain first** — facts once, unambiguous, annotated, current. 2. **Then the profiles** — map service, directories, trade portals, all on the same state and pointing back to the source. 3. **Then visibility elsewhere** — specialist articles, mentions, reviews. They work because there is something to point at. The other way round does not work. Not because the order is an opinion, but because steps 2 and 3 refer back to step 1 — and a reference to something unclear makes the confusion bigger, not smaller. ## What this has to do with ownership A platform can change its rules, throttle its reach or disappear. Your domain cannot, as long as it belongs to you. That is not a romantic argument but a practical one: the foundation is the only part of your digital presence where you decide what it says and how long it stays. That is why the question "who owns the place where my facts live?" is not a side question. It is the first one. ## The honest limit A clean source makes sure the facts are right when someone retrieves them. It does not make anyone ask. Whether your business gets recommended is decided by what you offer and what others say about you — not by the structure of your website. The foundation makes you citable, not popular. Confusing the two is the most common mistake in this field, and it is readily sold. And whether it works belongs to measurement rather than belief: [how you can tell](https://btlabs.dev/en/posts/what-you-measure-when-position-one-is-worth-nothing) is covered in a separate article. How many places actually diverge in your case is shown by the [route from the island zoo to a single source](https://btlabs.dev/en/posts/from-island-zoo-to-source-of-truth). If you want to know where your source sits — and whether it deserves the name —, [talk to us](https://btlabs.dev/en/contact). A short plain-talk conversation is enough for an honest assessment. ## Sources - Tow Center for Digital Journalism, Columbia University (March 2025), "AI Search Has a Citation Problem": 8 tools, 1,600 queries, more than 60 % incorrect source attributions; spread from 37 % to 94 % depending on the system --- Source: https://btlabs.dev/en/posts/your-domain-is-the-source Last-Modified: 2026-09-10T07:00:01.194Z Languages: [de](https://btlabs.dev/llms/de/posts/deine-domain-ist-die-quelle) · [it](https://btlabs.dev/llms/it/posts/il-tuo-dominio-e-la-fonte) See also: [llms.txt](https://btlabs.dev/llms.txt) · [ai.txt (Policy)](https://btlabs.dev/ai.txt) · [identity.json](https://btlabs.dev/identity.json)