Why traditional wire services don't get cited by ChatGPT

For seventy years, the press release wire was a closed loop. You paid PR Newswire or Business Wire $1,000+, your release went out to Bloomberg terminals, AP feeds, and Yahoo Finance, and "distribution" meant landing in the inboxes of human journalists.

That model is now invisible to the systems that decide what gets cited online.

PressGEO Research ran 1,100+ citation checks against the seven AI engines that matter — ChatGPT, Claude, Gemini, Perplexity, Copilot, Grok, and Reddit's answer surface. For every query about a recently announced product, funding round, or partnership, we tracked which sources the engines pulled from. Traditional wire services almost never appear.

There are four structural reasons — the wrong domain, the wrong HTML, the wrong distribution graph, and the wrong freshness model. We unpack each in the full analysis on the release page.

Read the full analysis →

The short version

• Domain: Your release lives at prnewswire.com/..., not on your own domain. AI engines see the wire as the source, not you.
• HTML: Big wires ship pages with no JSON-LD Article or Organization schema, no FAQPage markup, no llms.txt. Retrieval systems weight structured data heavily; the wires don't have it.
• Distribution: "10,000+ media outlets" is mostly low-signal aggregators that AI engines have learned to ignore. One well-structured release on your own domain out-cites a $1,200 wire blast.
• Freshness: Wires treat releases as 24-hour events. AI engines treat them as durable knowledge that needs dateModified and ongoing signal.

The playbook that does get cited is the opposite of all four: publish on your own domain, wrap it in Article + Organization + FAQPage schema, expose a machine-readable feed plus llms.txt, and reference the release from at least one secondary surface.

That's what PressGEO is built for. Read the full release.