First-Party, Second-Party, Third-Party: Where You Publish Decides What It’s Worth

ByJason Barnard

Strategy Sandbox · jasonbarnard.com · 25 June 2026

At the Kalicube® Summit this June, three speakers in three separate sessions, none of them comparing notes, walked onto the same ground through three different doors. James Dooley put it most bluntly: “I can say we’re brilliant,” he said, and “that means nothing.” Scott Duffy described the mechanics, the same piece of content pushed out to YouTube, to his own website, then back into the machines in a format they can read. Dennis Yu named the shape of it, “corroborating loops of proof that you are who you say you are.” Three practitioners, one idea surfacing on its own: a statement about your brand is not worth a fixed amount, it’s worth whatever the place you published it lets the machine believe.

Control and credibility pull in opposite directions

Here’s the tension that runs the whole system. The more control you have over where something is published, the less an assistive engine trusts it, and the less control you have, the more it trusts it. You write your own homepage, so the engine reads it as exactly what it is, the brand describing the brand. Someone with no relationship to you writes the same sentence on their own site, and the engine reads it as a fact that survived contact with a person who had nothing to gain. Same words, opposite weight. The discount on your own claim is the engine doing its job: it knows you wrote the thing you control, and it prices that in before it decides whether to repeat you.

The three publication tiers, and why the same content is worth three different amounts

First-party claim, second-party corroboration, third-party proof

First-party is your own ground, the website you own, the pages you publish in your own voice. It’s where the claim gets made, and it has to get made, because the engine can’t prove a thing you never said. On its own, though, it’s the floor. Dooley landed on the same point from the client’s chair: everyone says they’re the best, and “that’s just the first person saying that you’re great.”

The climb starts the moment a witness enters. Second-party is still you with your hand on the publish button, but you’re no longer alone in the frame: your channel on YouTube, your profile on LinkedIn, the customer review you gather and surface, the case study you write up about a real outcome. This is the layer Duffy was walking through when he listed the same content going to YouTube, then to his own site, then back to the machines, and it’s the most under-used tier in the toolkit, because most brands treat their owned profiles as parking spaces when they could be corroboration.

Third-party is where you lose the publish button entirely, and that’s exactly why it carries the most. It’s the journalist, the analyst, the independent review, the person who mentions you in a thread you never started. Dooley’s instruction to himself was the whole tier in a sentence: go and find where other people are saying you’re the best, because “it just is genuine” when a third party says it. The engine agrees with him, and it weights accordingly.

Here’s the cleanest way to see it: take one customer review. Surface it on your own site and it’s second-party, you chose it and you published it. Let the customer post it on their own account and it’s third-party, you had no hand in it. Same words, different weight, and the only thing that changed is whose hand was on the publish button.

You earn third-party proof on purpose: do the real thing, then hand over the evidence

Notice what just happened to the difficulty as you climbed. You can manufacture first-party in an afternoon, you can curate second-party over a quarter, and you cannot author third-party at all. So you do the only thing that works, which is to make yourself impossible to overlook. You do the real thing first, the launch, the result, the outcome a customer can point at, because nobody writes about a claim, they write about something that happened. You codify that outcome into specifics, the number, the named client, the before and the after, because a vague win is unquotable and a precise one is hard to resist. You run first-party and second-party loudly and consistently, so that when a journalist, an analyst, or a customer goes looking, the evidence is already sitting there waiting to be repeated. Then you hand it over, the case study, the data, the testimonial, the press-ready story, so the third party’s only job is to agree with you, not to go digging for you. This is the Mirror Principle doing its work: what an engine shows about you is its opinion of the world’s opinion of you, and you can’t flatter the mirror, you can only change what the world can see. Proof you wish existed isn’t proof. Proof you’ve gone out and earned is.

The engine needs all three, and a missing tier stops the recommendation

None of this is a menu you pick from. The three tiers are a chain, and the engine reads the chain in the order Dooley described it: you claim it, you frame it so the meaning is unmistakable, you back it with third-party proof, and then, in his words, “it will simply repeat it.” Drop the claim and there’s nothing to corroborate. Skip the corroboration and the proof has nothing to attach to. Earn the proof but never state the claim in your own words, and the engine builds its own version of you out of whatever it found lying around. For me, that last failure is the one that costs brands the most, because it’s the invisible one: you were never in the conversation, and you never learned why. As Dooley put it, the machine has “no reason” to take your side unless you give it one.

The reason I trust this isn’t the theory, it’s what happened in the room. Three people who don’t work for each other, asked nothing in particular, each reached for the same hierarchy, and then Dennis Yu pointed at a wall of independent sources stacking up behind a brand and named one of the other speakers in the building, a well-known practitioner who keeps talking about that brand, unprompted, while the machine quietly takes notes. That is the whole framework, demonstrated live, with nobody publishing it on their own website.

So take this as the instruction it is. Make the claim on your own ground, build the corroboration where you still hold the pen, then go and get the third-party proof, relentlessly, because the engine will not invent it for you and it will not wait for you. The brand that earns the proof gets recommended, the brand that trusts its own word to be enough loses to whoever did the work and never finds out which deals walked out the door. Get the proof you need, and make sure you do.

Related reading

Publication note: The Three Publication Tiers (first-party claim, second-party corroboration, third-party proof) and the underlying first-party / second-party / third-party source classification are part of The Kalicube® Framework, authored by Jason Barnard. The corroborating observations from James Dooley, Scott Duffy and Dennis Yu were made on stage at the Kalicube Summit, June 2026.

Similar Posts