A brand is a provable claim of origin long before it is an image. This article connects the history of the cattle brand with the present of answer engines: ownerless brand campaigns, the Graph Loop as a self-reinforcing cycle, the standardization arc from Schema.org to C2PA, and the work by which an entity makes its provenance machine-verifiable.
There is a moment just before sunrise when the prairie belongs to no one. No fence, no land registry, no sheriff. Just open country and whatever stands on it. Anyone who owned cattle in the American West of the nineteenth century knew that moment well. Ownership out there was a mark, a glowing iron pressed into the flank. The animal roamed freely across open land, yet the mark said unmistakably who it belonged to.
That image comes to mind every time I scroll through my feed these days. For weeks now, talented people have been publishing campaigns for Nike, for Adidas, for some of the most closely guarded brands in the world. High gloss, sharp headlines, visual language at agency level. It looks like the real thing. Nobody commissioned it, and nobody at the brand ever saw or approved it. These are concept pieces, pitches without an addressee, brilliant finger exercises on someone else's logo.
I understand the appeal completely. This is skill looking for a stage, the desire to play with the big names for once instead of just talking about them. It is human and it is brave, and some of this work is better crafted than what the real brands actually put out. But this is exactly where the problem begins. And it is a different problem than most people assume.
A brand was never an image
Criticism of this wave almost always runs through taste or law. Whether you are allowed to do it, whether it is trademark abuse, whether it damages your own credibility. All legitimate questions, and all secondary. The real problem is technical, and it has a name that never comes up in any LinkedIn discussion: ownerless provenance.
A quick look at the history of the word settles the matter in one stroke. Brand comes literally from the branding iron. On the open range, with no fence and no register, that mark was the only means of making ownership provable at all. It was a certificate of origin, proof that this animal came from this ranch. Anyone who caught someone else's cattle and rebranded it was a rustler.
From this follows the thesis that carries this text:
A brand is a provable claim of origin long before it is an image.
Once you internalize that, you see the wave with different eyes. What matters then is not so much whether the campaign is beautiful. What matters is whether its provenance is true. And it is not. To stay with the image, this is rebranded cattle. Someone else's animal with a famous mark pressed into its flank. It runs beautifully, it runs convincingly, and the origin it claims is false.
The machine is the new open range
To understand why this weighs heavier today than ever before, you have to know how a machine comprehends a brand. An answer engine does not know brands. It knows entities: uniquely identifiable things to which it attaches attributes, relationships and evidence. It builds that entity from thousands of distributed signals across the entire web. Every mention, every link, every structured data field is a ballot on what the entity is and who it belongs to.
That is the new open range. Open country, no fence, no land registry. The machine walks across it and reads the marks. And it does not read the poster. The raw pixels of a beautiful layout remain mute to it. It reads the language around the image: the caption, the alt text, the hashtags, the comments, the reposts, the articles, the links. An entity emerges from textual co-occurrence and linkage. If a brand's name appears often enough together with certain statements, a creator and plenty of resonance, that is a signal, regardless of whether the work was ever authorized. And the gap keeps closing, because modern systems increasingly pull marks out of the image itself through OCR and logo recognition.
In between sits a layer that classical communication never pays attention to and that decides everything in AEO: the entity layer. The structured level on which a brand explicitly tells the machine who it is, where its statements come from and which authoritative sources it is identical with. Whoever fails to shape this layer actively leaves their entity's image to the chance of the open range. And today, chance also means: ownerless campaigns nobody ever ordered.
The Graph Loop that feeds itself
These signals do not behave statically. They run in a circle. An authoritative source makes a statement about the entity. Others pick it up and confirm it. The knowledge graph consolidates the agreement into a fact. The answer engine serves the fact, and that serving generates new content that in turn confirms what has already been served. A closed, self-reinforcing cycle. I call it the Graph Loop.
The loop is magnificent when it amplifies clean signals, and dangerous when it amplifies noise. It does not distinguish between true and false, it amplifies whatever appears frequently and consistently.
Here we have to stay honest, or the argument tips into melodrama. At Nike, practically nothing happens. Nike is one of the most heavily attested entities in the world, a dense web of confirmation built from millions of signals. A single ownerless pitch is noise far below any threshold. The danger runs exactly the other way: it scales inversely with corroboration strength. The thinner an entity's web of confirmation, the more easily misattributed signals move the needle. That hits small and midsize brands, niche topics and, this is the real point, the professional entity of the creator themselves. Whoever brands ownerless work for Nike dilutes above all their own, still weakly attested entity by loading it with foreign provenance.
So the wave cuts two ways. One edge is real and good. It shows talent, trains craft, makes people visible, and occasionally brands discover new voices exactly this way. The other edge works more quietly. It feeds ownerless signals into the loop, charges the more weakly attested entities and does so at the very moment machines are beginning to evaluate provenance. What was a harmless portfolio piece yesterday is a signal tomorrow in a system that does not forget.
We have tamed the West once before
None of this is new. We have stood exactly here before. In the early years of the structured web, a dialect war was raging. Every major search engine favored its own format for machine-readable data. The cattle wore a different brand depending on the pasture, and nobody could be sure the neighbors could even decipher it.
Then something rare happened. On June 2, 2011, rival search engines, Google, Bing and Yahoo, sat down at one table and founded Schema.org, a shared vocabulary for structured data. It was a peace treaty that standardized the brands. What happened next is what most people overlook. Competition did not disappear, it moved upward. Once reading the web became shared, neutral infrastructure, the fight shifted to the question of what you do with a readable web. The brand mark went from advantage to price of entry. Whoever spoke no schema simply existed less for the machine.
On that pacified range, the next structures rose: rich results, the Knowledge Graph, the knowledge panels, voice search and, at the end of that chain, the answer engines we are talking about today. The lesson is unambiguous: order always comes. The only question is who holds the branding iron when it arrives.
And it is coming again, just one floor deeper. In 2011 the issue was meaning: making machine-readable what a thing is. Today the issue is provenance: making machine-verifiable where something comes from and who made it. The Coalition for Content Provenance and Authenticity, C2PA for short, founded in 2021 by Adobe, Arm, BBC, Intel and Microsoft, now counts thousands of members, among them Google, Meta, OpenAI and Sony. In 2025, its specification became a formal ISO standard. This time a sheriff is writing along who did not exist in 2011: the legislator. With Article 50 of the AI Act, the EU requires machine-readable labeling for AI-generated content starting August 2026.
The treaty is signed, but the territory is not yet pacified. Social platforms strip embedded provenance data on upload, so the chain breaks exactly where reach is created. That is precisely the hole the ownerless campaigns ride through. Beautifully branded cattle whose origin nobody checks, because the verification infrastructure is still full of holes.
Why none of it converts
There is a telltale finding that anyone who looks closely can confirm. These campaigns generate applause but no conversion. Admiration in the comments, zero revenue anywhere.
From the provenance perspective, that is an inevitable consequence. Conversion arises from a verifiable path from a signal to an action. That path needs a destination with registered provenance, an owner to whom the machine can route an action in the first place. This is exactly what the ownerless campaign lacks. No Entity Home to anchor it, no corroboration to confirm it, no structured bridge between the content and a real transaction path. It is applause without an entry in the land registry, and it evaporates because it never arrives anywhere.
The work of the Entity Architect
What follows from this in practice, for every brand and every professional who takes their own entity seriously? The same answer as in 2011. Back then, the winners were those who registered their brand mark before the machines hardened their view of the world. That is repeating now, in a clear sequence.
First, establish the Entity Home: the one canonical source where the entity fixes its provenance authoritatively. The registered brand mark, one place everything else points to.
Second, make that place machine-readable. Structured data, a cleanly marked-up organization and, above all, explicit sameAs connections to the authoritative profiles that attest the same entity elsewhere. This is the language all machines have understood since 2011, and today it is the price of entry.
Third, build corroboration. A statement about your own entity only becomes a fact when independent, authoritative sources confirm it consistently. These are the witnesses on the range who certify that the mark is genuine. A single claim remains a claim; an attested claim becomes truth in the graph.
Fourth, close the Graph Loop deliberately. When Entity Home, structured data and corroboration agree, the cycle amplifies the clean version of the entity instead of the noise. Whoever does not close it themselves leaves it to others, in doubt to yesterday's pretty, ownerless campaign.
That is the core of what I call Machine First. A sequence. First make sure the machine knows the provenance beyond doubt, then fight for human attention. Whoever works this way builds on the land registry. Whoever reverses the sequence builds on open range.
The mark, before it is too late
One sentence from the current provenance debate sums up the whole situation: the absence of a proof of origin is slowly becoming a signal in itself. For a long time, ownerless content was neutral; nobody asked where it came from. Once machines check provenance routinely, content without provable origin becomes suspect. The rebranded animal then stands out through the missing, genuine register behind it.
The Wild West is back, yes. But the West was never the land without marks. It was the land where the mark decided everything, precisely because there was no other order. Those who understood that branded their iron early and deep.
Whether order is coming is not up for debate. It comes every time, and faster than most people believe. The only open question is whether your entity is already in the land registry while everyone else is still fighting over land. That is exactly the work we do at richresults.ai: we set up the Entity Home, make provenance attestable and close the Graph Loop, so that AI systems find an organization understandable, citable and recommendable. That is Entity Building, and it starts before the rest of the range has even noticed that the surveying has begun.