When I Choose Distribution Over a Logo

I do not choose the logo a board would recognize, and I do not choose the outlet that can spray the story across the most domains. On 16,039 monitored answer runs, neither prestige nor distribution volume wins by itself. forbes.com ranks second of 1,230 classified editorial publications. prnewswire.com ranks eighth. Forbes is cited 3.65 times as often as the wire, so volume is not the winner. Reuters, one of the most trusted newsrooms there is, does not clear the evidence floor, so prestige is not the winner either. What separates the top of the table is recurrence: engines coming back, on many days, on every engine.

The measurement is Which Publications AI Search Engines Cite Most in 2026. That page is the deeper dataset. This essay is how I would use it as a founder: the column I use when a logo and a wire are both on the table, the scope of that choice, and the reading that would put the logo back in front.

The numbers the choice has to survive

Machine Relations Index release mri_score_v2.0+2026-09-20+47973f373a20, window 2026-05-10 to 2026-09-20, 127 days, 919 monitored prompts, 16,039 answer runs, 22,320 cited domains, six engines. A domain counts once per answer run in which it is cited. A rate publishes only past 10 observed runs across 7 distinct run dates. Below that floor the release says Collecting, and I leave it there.

Publication	Citation rate	Cited runs	Days cited	Engines	Confidence	Standing
medium.com	5.04%	808	109 of 127	6	A	rank 1 of 1,230
forbes.com	4.15%	665	106 of 127	6	A	rank 2 of 1,230
prnewswire.com	1.13%	182	75 of 127	6	B	rank 8 of 1,230
techcrunch.com	1.02%	164	44 of 127	5	B	rank 9 of 1,230
fortune.com	0.37%	59	28 of 127	4	C	rank 36 of 1,230
reuters.com	0.08%	13	13 of 127	3	Collecting	below the floor for a rank

The class rank is not a clean newsroom ranking, and I do not pretend it is. source_role in this release combines source-type evidence with AuthorityTech's placement catalog. 96 of 1,231 class members enter through the catalog. The class includes six platform domains with the 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 are filed as community platforms even though they are the same kind of hosted publishing product as Medium. Those six platform domains hold 1,358 of the editorial class's 13,525 cited runs, 10.04%, and they include the top slot. The class denominator moves. Published figures run from 1,025 to 1,230. Rate, cited runs, days cited, engine breadth, and confidence do not move with that relabeling. I decide from those.

Why a count hands the decision to the wire

A real 30-day window goes up and down. A number that only climbs is counting since the count began, and it measures how many URLs a source has put into the corpus. A wire publishes one release to hundreds of syndicated URLs. A newsroom publishes one story to one URL. Count URLs and the wire wins by construction. Count answer runs and it does not. Any figure that rises every week under a "rolling window" label gets restated as a rate before I will use it in a meeting.

Reuters makes the prestige failure specific. It is cited in 13 of 16,039 runs, on 13 days, across 3 engines, below the floor at which this instrument will publish a rank. Fortune sits at rank 36. The masthead in the boardroom is not the source in the answer. I can admire the newsroom and still refuse to pay a prestige premium for a citation job the release does not show it doing.

Recurrence is the column I buy

The days-cited column is the one a pickup report cannot show. Medium is cited on 109 of 127 observed days and Forbes on 106, both across all six engines. PR Newswire appears on 75 days. TechCrunch on 44. Reuters on 13. The top of the table is not the biggest publisher. It is the source engines return to, across questions, on different dates, on every engine.

The wire class is the direct test of the distribution thesis. Nine domains in wire and press-release distribution take 232 cited runs out of 112,518 across the nine classes, 0.21% of classified citation. Across 88 published segments on the 2026-09-19 release, wire distribution held 0 of 880 top-ten slots. Maximum syndication, almost no selection.

Two figures in that paragraph look like they fight the rank-8 row, so I keep the reconciliation in the decision and not in a footnote. prnewswire.com is classified as an editorial publication in this release, not as wire and press-release distribution, so the nine-domain class excludes it. Counted with the wires, the class is ten domains holding 414 of 112,518 cited runs, 0.37% of classified citation. prnewswire.com's best standing in any published segment is 15 of 448. The class still holds 0 of 880 top-ten slots. Rank 8 of 1,230 and the class share sit on two denominators. I do not average them into a story where the wire both won and lost.

The decision

When the goal is an answer-engine citation, I choose the source with recurrence: a high count of days cited, on as many of the six engines as the release shows, at a confidence grade I can say out loud. On this release that is Medium and Forbes before a prestige name that sits at rank 36 or below the floor, and before a wire whose class does not appear in the top ten of a buyer segment. I read days cited before I read the rate. A source on 100 of 127 days is being selected. A source on 20 days is catching bursts. PR Newswire's 75 days is real recurrence at a 1.13% rate. I still do not buy it as the citation line, because the class test and the top-ten test fail.

I treat a wire release as distribution to the audiences it was built for, and I price it that way. I ask any vendor quoting a citation number for the window, the denominator, and the evidence floor. A count with no denominator is inventory. The engines publish their crawlers, which is the standard of specificity I want mirrored in the citation claim: Google's common crawlers, OpenAI's bots, Perplexity's bots, and Anthropic's crawler policy.

The vendors are selling the distribution version of the argument in their own words. GlobeNewswire, read 2026-09-20, offers "the blueprint for AI-visible press releases." globenewswire.com is cited in 74 of 16,039 monitored runs, at confidence C. The positioning and the standing are separate facts. The only defense I have against buying the first as if it were the second is an instrument that publishes its denominator.

The scope is the choice between a recognizable logo, a high-volume wire, and a source with recurrence, for citation spend, on this release. It does not govern a story whose job is a human reader, a regulator, or a customer who asked for a specific masthead. Those jobs can still be worth the logo. They are not this measurement.

I would put the logo back in front of recurrence if a later release showed Reuters, or another prestige newsroom now below the floor, clearing 10 runs across 7 dates with a rate and a days-cited figure in the range of the current leaders, above 4% and on roughly 100 of the window's days. I would put distribution volume back in front if the wire class, with PR Newswire counted in, cleared a meaningful share of classified citation and took top-ten segment slots instead of 0 of 880. Neither is true on the 2026-09-20 release. Until one is, recurrence outranks the logo and the pickup count.

Machine Relations is the frame I use so distribution stays a layer and does not get promoted into the whole discipline. The class names and the rate are in the glossary. A discipline whose numbers I cannot re-open is a brochure. This release id is the thing I can re-open.

FAQ

When do I actually pick the logo?

When the audience is a person who uses the masthead as a trust shortcut, and the budget line is that person's attention. Fortune at rank 36, 59 of 16,039 runs, confidence C, can still be the right human placement. I do not then report those 59 runs as if they were Medium's 808. The logo wins a different contest. I keep the contests in different sentences.

When do I pick distribution volume?

When I need the story in the databases and on the surfaces a wire was built to fill, and I am willing to pay for reach rather than for selection. Nine wire domains hold 232 of 112,518 classified citation runs. Adding PR Newswire makes it 414, still 0.37%, still none of 880 top-ten slots. I buy the reach with that ceiling written on the purchase order.

Why does Medium beat a famous newsroom?

Because days cited and rate are the measures. Medium is on 109 of 127 days, six engines, 5.04%, confidence A. Reuters is on 13 days, three engines, 13 runs, below the floor. I am not claiming Medium is the better newspaper. I am claiming it is the source this window shows engines returning to. A founder who needs the newspaper for a human reason should buy the newspaper and call it that.

What would flip the choice?

A prestige domain clearing the evidence floor with a rate near the current leaders, on the order of the 4% and the hundred-day recurrence Forbes and Medium show, or the wire class taking top-ten slots in published segments. GlobeNewswire's own line, 74 of 16,039 at confidence C, is not that flip. A homepage that calls the release a blueprint for AI-visible press releases is positioning. I would need the rate.