Why I Separate Wires from Citation

I price a wire release as distribution to the audiences and the disclosure jobs it already has, and I price citation from a different table. Measured on one denominator, 16,039 monitored answer runs, Medium ranks first among the domains this release calls editorial publications, Forbes second, then TechRadar and Yahoo. PR Newswire ranks eighth. Forbes is cited 3.65 times as often as PR Newswire. I would not let a wire invoice and a citation strategy share a line, because the counts do not share a magnitude.

The publication table I am separating the wire from is Which Publications AI Search Engines Cite Most in 2026. That page is the deeper resource, the dataset. This essay is how I would use it as a founder: where I would still buy a wire, why the masthead is the wrong first question, and what share would make me put wires back on the citation line.

The table I would keep in two budgets

Release mri_score_v2.0+2026-09-20+47973f373a20. Window May 10 to September 20, 2026, 127 days, 919 monitored prompts, 16,039 answer runs, 22,320 cited domains, six engines: ChatGPT, Claude, Gemini, Google AI Mode, Google AI Overviews, and Perplexity. A domain counts once per answer run in which it is cited. A rate publishes only after 10 observed runs across 7 distinct dates, and a confidence grade travels with it.

Publication	Citation rate	Cited runs	Days cited	Engines	Confidence	Standing
medium.com	5.04%	808 of 16,039	109 of 127	6	A	1 of 1,230 classified editorial publications
forbes.com	4.15%	665 of 16,039	106 of 127	6	A	2 of 1,230
techradar.com	2.87%	460 of 16,039	not published on this row	4	A	3 of 1,230
yahoo.com	1.93%	310 of 16,039	not published on this row	5	A	4 of 1,230
prnewswire.com	1.13%	182 of 16,039	75 of 127	6	B	8 of 1,230
techcrunch.com	1.02%	164 of 16,039	44 of 127	5	B	9 of 1,230

The editorial-class rank needs a label before anyone uses it to buy a wire or a host. It is read off the Index's source-role field, which in this release combines source-type evidence with a placement catalog. Of 1,231 class members, 96 are admitted through the catalog, and editorial classification is moving to source-type evidence alone. The class includes six platform domains beside newsrooms: medium.com at class rank 1, amazon.com at 7, prnewswire.com at 8, apple.com at 16, wordpress.com at 239, and blogspot.com at 330. Substack, Beehiiv, and Dev.to, the same hosted open-publishing product as Medium, are filed as community platforms. Those six hold 1,358 of the editorial class's 13,525 cited runs, 10.04%, including the top slot. The class denominator moves between releases, with published figures from 1,025 to 1,230, because it counts the domains classified each time. Rate, runs, days, engines, and confidence are direct observations. I would buy from those. I would not buy from the noun "editorial" while PR Newswire and Medium are inside it for different reasons.

The wire and press-release distribution class, as a class, is 9 domains holding 232 cited runs out of 112,518 across the nine source classes, 0.21% of classified citation. The Index files prnewswire.com, the largest wire domain in it, as an editorial publication, so that nine-domain class excludes it. Counted with the wires, ten domains hold 414 of 112,518 cited runs, 0.37%. PR Newswire's best standing in any published segment is number 15 of 448, so the class still holds 0 of 880 top-ten slots. Business Wire is cited in 98 of 16,039 runs at confidence C. GlobeNewswire is cited in 74 at confidence C. Across all 88 published segments on the September 19, 2026 release, wire distribution held none of the 880 top-ten slots.

I would put 5.04%, 4.15%, 2.87%, and 1.93% in the citation budget, with the class caveat on Medium. I would put 0.21% and 0.37% in the distribution budget, as the reason a wire is not a citation strategy. Forbes at 665 runs against PR Newswire at 182 is the 3.65 times I would say out loud if someone calls the wire a substitute for a newsroom.

The masthead is the wrong first question

Forbes at rank 2 of 1,230 means prestige and citation can travel together. They do not travel together automatically. The four outlets founders are often sold as one interchangeable tier span a 59-fold range of measured citation rate, and inside a single publication the question shape a piece answers moves its citation rate by roughly 31 times. Both figures are measured, and both are in the peer analysis of how a feature actually gets cited. I would not start with "how do I get into Forbes." I would start with whether the page answers the question a buyer asks, in a form an engine can lift. A famous masthead with the wrong shape can lose to a plainer page with the right one. A wire release that multiplies URLs does not answer that question at all.

That is the operational content of the discipline I describe on Machine Relations. The authority I would require the page to carry, rather than assume from the logo, is earned authority.

What the wire vendors are selling beside the count

I would read the sales language next to the standing, and I would not average them. GlobeNewswire, read on September 20, 2026, offers to help shape a presence in AI answers and leads with a framework it calls the blueprint for AI-visible press releases. PR Newswire carries a Create with AI product. Cision has announced the acquisition of a generative-AI monitoring platform. GlobeNewswire's domain is the 74 of 16,039 runs, on 37 of 127 days, at confidence C. The positioning is a sentence on a website. The standing is a count. Only the count is something I can check.

I would still buy a wire when a financing, a filing, or a customer announcement needs to reach the wire's original audiences on a clock. That purchase has a price and a job. Calling the same invoice an AI citation program takes a 0.21% class and prices it like a 5% domain. I would rather show both numbers than let the blueprint language replace them.

What I would do with the table

I would rank citation targets by measured rate on one denominator, with the window, the run count, and the confidence grade on the row. Every publication in the Index carries those fields. I would keep wire distribution on the obligations it has always covered. I would ask any vendor who quotes a citation number for the window, the denominator, and the evidence floor, and I would omit the number if I cannot get them. I would make our own pages worth citing, because the crawlers are documented by the engines: Google, OpenAI, Perplexity, and Anthropic. A page those crawlers can read, which answers a buyer question, is the asset the 31-times shape effect says is worth more than the logo.

Pew's account of how Americans view AI is why I treat the demand side as structural. A structural demand is a reason to keep an instrument, the 16,039-run window, instead of a cumulative counter that rewards syndication.

The decision

I would split the next planning doc into a citation list and a wire list. The citation list is sorted by rate on this release: the domains at 5.04%, 4.15%, 2.87%, and 1.93% before the domain at 1.13%, and TechCrunch's 1.02% and 44 days as a thinner selection than its reputation suggests. The wire list is priced for disclosure and reach, with Business Wire's 98 runs and GlobeNewswire's 74 runs attached so nobody confuses a send with a citation. I would also refuse to treat the editorial-class rank as permission to call a wire, or a host, a newsroom.

Scope: budget separation on release mri_score_v2.0+2026-09-20+47973f373a20 and the September 19 segment slot counts. It covers the six engines named above. It does not say a wire is useless, and it does not say TechRadar's missing days-cited cell can be filled in from memory. Where the row does not publish days, I would leave the cell blank rather than borrow Forbes's 106.

What would change my mind: a later release in which the wire class, on a denominator I can see, held top-ten segment slots and a share of classified citation that a founder could confuse with Medium's 5.04% or Forbes's 4.15%. I would then put wires on the citation list and say why. I would also collapse the two budgets if the 59-fold tier gap and the 31-times question-shape gap disappeared in the peer measurement, because those are the reasons a masthead and a wire are both the wrong units. I would not collapse them because a vendor's blueprint used the words "AI-visible."

FAQ

Why is Forbes 3.65 times PR Newswire, and why is that not the whole argument?

Forbes has 665 cited runs and a 4.15% rate. PR Newswire has 182 runs and a 1.13% rate. 665 divided by 182 is the 3.65. The wider argument is the class: nine wire domains at 232 of 112,518 runs, 0.21%, or ten domains at 414 runs, 0.37%, with 0 of 880 top-ten slots. PR Newswire's best segment rank is 15 of 448. I would use the multiple to answer "isn't the wire basically Forbes." I would use the class share to answer "isn't the wire system how AI citation works."

Where do TechRadar and Yahoo fit?

On this release TechRadar is third in the editorial class, 2.87%, 460 of 16,039 runs, four engines, confidence A. Yahoo is fourth, 1.93%, 310 runs, five engines, confidence A. Days cited are not on those two rows in the table I am using, so I would not invent them. Both outrun PR Newswire's 182 runs. I would pursue them as citation targets when the page shape matches a buyer question, and I would not assume the class rank means a staff editor chose a contributed post.

What would I still pay a wire for?

The audiences and the obligations that existed before anyone sold an AI blueprint: a disclosure, a launch that must be on the wire, a distribution list the wire actually has. I would pay that price on that line. GlobeNewswire's 74 of 16,039 runs, 37 of 127 days, confidence C, is the standing I would show next to the September 20, 2026 blueprint language. Cision's monitoring acquisition and PR Newswire's Create with AI product are product facts. They are not a change in the 0.21%.

What makes a page inside a prestigious outlet worth more than the outlet?

The question it answers. A 59-fold range separates outlets that are sold as one tier, and question shape inside one publication moves citation rate by about 31 times, in the peer measurement I cited. I would brief the writer on the buyer question and on a sentence an engine can lift. I would not brief the writer on the logo and hope the 31-times effect goes our way by accident.