Key takeaways
- Press releases and news are a small slice of what ChatGPT cites. In Promptwatch Data for Sep 28 to Oct 4, 2026, product pages (22.4%) and landing pages (16.8%) lead, while news is 5.7%.
- Third-party listicles (13.6%), comparisons (5.9%) and reviews matter more to AI answers than most PR teams assumed.
- News coverage decays fast in ChatGPT: news pages lose half their citations in about 36 days, while landing pages and docs last for months.
- Each AI engine cites social platforms very differently. YouTube is 6.71% of Google AI Overviews citations, while social is only 0.64% of ChatGPT's.
- Comms teams are adding AI visibility, sentiment and citation tracking next to clip counts and share of voice.
Why PR teams suddenly care about AI search
For years the PR scorecard was clips, reach, share of voice and the occasional backlink. Those numbers still matter, but they do not tell you what ChatGPT, Gemini or Perplexity says about your company when a buyer, journalist or investor asks.
That is the shift. AI answers are now a place where reputation gets summarized, and the summary is built from sources you may not control. Muck Rack's February 2026 webinar, "How AI affects PR in 2026", framed it as comms teams being asked to think beyond traditional search and media metrics, and pitched its Generative Pulse product as a way to see influence across AI platforms. The session summary contains no hard figures, so treat it as a signal of where the vendor market is heading, not as evidence of how big the shift is.
For numbers, you need citation data. That is where the rest of this guide spends its time.
What AI engines actually cite (and what that means for PR)
The most useful finding for comms teams is that earned news is not what drives most AI citations. Promptwatch publishes first-party research on this, and its ChatGPT citation share by content type report (Sep 28 to Oct 4, 2026) breaks it down.

| Content type | Share of ChatGPT citations |
|---|---|
| Product pages | 22.4% |
| Landing pages | 16.8% |
| Listicles | 13.6% |
| How-to guides | 8.7% |
| Documentation | 7.5% |
| General articles | 6.1% |
| Comparisons | 5.9% |
| News | 5.7% |
| FAQ pages | 3.3% |
| Reviews | 1.7% |
Two things stand out. First, 39% of citations go to pages that describe what a company sells. Your own website is doing a lot of the work. Second, news is under 6%. A big media hit is nice, but it is not the main route into a ChatGPT answer.
Branded prompts are different
When someone asks about a specific brand, the mix shifts in ways PR should care about. On branded prompts, product pages are 23.3% and landing pages 22.8%, but reviews climb to 5.6%, more than three times their share across all prompts. Press releases reach 2.1%, higher than their overall share, and social posts are 3.0%.
So for reputation questions ("Is Company X any good?"), third-party reviews carry real weight, and press releases get a small bump. Not nothing, but not a strategy by themselves.
Commercial prompts reward listicles and comparisons
For commercial prompts, product pages (26.0%), landing pages (20.6%), listicles (15.0%) and comparisons (7.9%) add up to almost 70% of citations. If your category has "best X tools" roundups and you are not in them, that is a PR problem as much as an SEO one. Getting included in the listicles AI already cites is a very old PR task (pitching editors) pointed at a new target.
News has a short shelf life
The content half-life report tracked every page ChatGPT cited on July 6, 2026 for 90 days. News pages lost half their citations within 36 days. Landing pages and docs stayed cited much longer. A launch announcement can get you a quick bump, but the durable asset is a page you own and keep updated.
Social platforms: one rule does not fit all engines
PR teams love to say "be present everywhere." The data says be specific. Promptwatch's social media citations report (Oct 11 to Nov 9, 2026) shows how little ChatGPT and how much Google AI Overviews rely on social.
| Engine | Social share of all citations | Notable platforms |
|---|---|---|
| Google AI Overviews | 13.16% | YouTube 6.71%, Facebook 2.34%, Reddit 2.10%, Instagram 1.02%, LinkedIn 0.78% |
| Perplexity | 1.94% | Reddit 0.82%, LinkedIn 0.66%, YouTube 0.31% |
| ChatGPT | 0.64% | Reddit 0.40%, LinkedIn 0.20% |
YouTube alone is more than the other six platforms combined in AI Overviews. For a comms team, that makes video (executive explainers, product walkthroughs, spokesperson interviews) a more direct lever for Google's AI answers than another LinkedIn post. LinkedIn actually fell in AI Overviews, from 1.55% to 0.78%.
A caution on the numbers: the October report covered 7 days and the November one covers 30, so the month-over-month changes are not like-for-like. The relative ranking between engines is the safer takeaway. Promptwatch's own advice from this data is to put YouTube first for Google, focus on Reddit and LinkedIn for Perplexity, not rely on social posts for ChatGPT, and keep X out of AI search plans.
Also worth knowing: Reddit's share of ChatGPT citations collapsed from about 4% to about 0.5% on August 14, 2026. Any playbook built on seeding Reddit threads for ChatGPT visibility aged badly overnight.
How comms teams are adapting
I see four changes that are practical rather than aspirational.
1. Measuring AI answers next to media coverage
Teams add a monthly AI visibility readout: which prompts mention the brand, how it is described, who else appears, and which sources get cited. The sources list is the PR-relevant part, because it tells you which publications and review sites shape the answer.
Tools in this space fall into a few groups.
| Tool type | Examples | What comms teams use it for |
|---|---|---|
| Platforms that track and act | Promptwatch | Visibility, sentiment, citations, crawler logs, content creation |
| Monitoring-focused trackers | Otterly.AI, Peec AI, Profound, Scrunch | Seeing mentions and citations across engines |
| SEO suites with AI add-ons | Semrush, Ahrefs Brand Radar | Adding AI tracking to existing SEO reporting |
| PR-native monitoring | Muck Rack Generative Pulse | Linking AI influence to media relations workflows |


For a longer list of options, the GEO software directory at bestgeosoftware.com is a good place to compare.
Where Promptwatch fits for comms: Brand Narrative pulls out the individual claims AI makes about your company ("expensive for small teams", "weak support") and files them under themes like Pricing or Reputation, on every plan. That is closer to what a reputation team wants than a single visibility score. On Business and above, Fact Check lets you mark a claim inaccurate and finds the pages on your site that state it, so you know where the wrong idea is coming from. You can try it with the free AI visibility report or the 7-day trial, and confirm current pricing at promptwatch.com/pricing.
2. Pitching for the pages AI already cites
Instead of chasing any coverage, teams pull the list of third-party URLs that AI cites for their category prompts and pitch those publishers. Given that listicles, comparisons and reviews outweigh news, a placement in a well-cited roundup can beat a one-day headline. Offsite mention tracking helps here, since AI often names a brand inside a third-party page without a link back.
3. Treating owned pages as PR assets
If product and landing pages are 39% of ChatGPT citations, the company website is the biggest PR channel nobody in comms owns. Comms teams are working with web and SEO teams on accurate, current pages: pricing, positioning, leadership bios, FAQs. They also check that AI crawlers can read the site at all. Tools that log AI crawler visits, including Promptwatch's AI crawler logs, separate an access problem from a content problem. The free Agentic Readiness Check is a quick first test.
4. Refreshing instead of announcing
Because news decays in about five weeks while landing pages and docs last, teams schedule refreshes of evergreen pages after each announcement, rather than letting the press release be the only record. Fact sheets, comparison pages and "about" pages get updated with new numbers.
A simple comms workflow for AI search
- Build a prompt list that reflects how journalists, customers and investors ask about you. Include branded, competitor comparison and category prompts.
- Run it across ChatGPT, Gemini, Perplexity and Google AI Overviews. Record mentions, position and the claims made.
- List the cited sources. Sort them into owned pages, earned media, reviews and listicles, and social.
- Fix owned pages first, since they are the biggest slice. Correct inaccurate claims at the source.
- Pitch the third-party pages that appear repeatedly. Prioritize listicles, comparisons and reviews over generic news.
- Match social effort to the engine: YouTube for Google, Reddit and LinkedIn for Perplexity, little for ChatGPT.
- Re-run monthly and compare. Report it alongside traditional coverage metrics.
If this sounds like a lot to run alongside normal media relations, that is why some brands bring in outside help. 1001 SEO Media is an agency that combines digital PR with generative engine optimization, and it uses Promptwatch for AI visibility work. Handy if your comms team lacks bandwidth, but the workflow above is doable in-house too.
Pitfalls to avoid
- Measuring one engine and generalizing. ChatGPT, Perplexity and AI Overviews cite very different source mixes.
- Overvaluing press releases. They are 2.1% of branded citations. Worth publishing, not worth building a strategy on.
- Chasing a single dramatic citation shift. Reddit's drop in ChatGPT shows how fast tactics can expire.
- Ignoring owned pages because they feel like "marketing's job."
- Treating a visibility score as the finish line. A score tells you where you stand, not what to change.
Promptwatch's Academy guide on citations vs mentions is a useful primer for teams new to these metrics.
What to do this quarter
Pick 25 prompts that matter to your reputation, run them across the main engines, and list the top cited sources. Then fix the three owned pages that get cited most, and pitch the three third-party pages you most want to be on. That gets you real data on how AI describes your brand, and a first set of PR actions that the citation numbers say are worth taking.

