Features
PR & media citations
The “Press relations and media citations” view starts from your AI observations: when an engine answers your tracked prompts, it cites sources. The third-party domains among them are the publications, directories and portals that already feed the answers your buyers read.
Where the data comes from
Each observed answer records the sources it cites, classed as pages of your site, a competitor's pages or third-party sources. The view reads the third-party sources of the last three completed runs of the site's prompt groups. Nothing is collected elsewhere: no media database, no press monitoring.
- Without a completed run: “No completed AI observation run for this site”.
- With runs but no third-party source cited: “No AI observation has cited a third-party source yet”.
- The figures change with every new observation: see AI prompt groups.
The KPIs
| KPI | Definition |
|---|---|
| Media cited | Distinct third-party domains cited in the answers read. |
| Citations | Citations of third-party sources, across all answers. |
| Verified citations | Citations whose link was checked online. That check is not performed yet: the figure shows “—” until a check has been made. |
| Answers citing a third party | Share of successful answers that cite at least one third-party source. |
| Domains without your site | Third-party domains cited in answers where your site is not cited: the opportunities. |
| Answers without your site | Answers citing a third party but not your site. |
| Link opportunities | Open link opportunities recorded by the backlink analysis. They are not press contacts: no outreach tracking exists in the product. |
The media list
The “Sources AI engines cite on your topics” card lists up to twelve domains, from most to least cited. For each:
- the number of citations and verified citations;
- Share of answers: for each engine, the share of its observed answers that cite this domain. It is a count of answers, not a trust score;
- the receiving engines that cited it;
- the domain's authority, only if a backlink reading already carries it; otherwise “No provider authority reading for this domain”.
The “Media by citations”, “Listed media cited, per engine” and “Observed AI answers” charts (third party without your site, third party and your site, no third-party source) show the same figures.
The AI press pitch generator
- Pick a target outlet among the cited domains, or “No specific outlet”.
- Optionally type your angle in “Email subject”.
- Click “Generate the AI pitch”: a draft email is written from the chosen outlet, your AI visibility figures and your site's topics.
- Review, edit, then “Copy the outreach email” and send it from your own mail client.
- Each generation is counted against your AI requests and needs a plan that includes them; a client viewer role cannot generate.
- To prevent bursts: 5 generations per minute per person, 30 per ten minutes per organisation.
- You can also write your message yourself in the fields, without AI.
How AI drafts work in general is described in AI help; the AI budget in Plans and billing.
How to use it
- Spot the domains without your site: an engine read a third-party source there instead of yours.
- Open the domain: which engine cites it, and on what share of its answers?
- Aim for a mention or a page on that outlet, directory or comparison site, and write your pitch.
- Run an observation again a few weeks later and see whether the answers cite you.
The overall method is in GEO; cited competitors in Competitor analysis.
What is not measured
- Media database: no catalogue of outlets, journalists or contacts; the “Sources required for measured press relations” card says so.
- Outreach: no CRM or send log, so no “contacted” status, reply, follow-up or response rate. The “Planned editorial pipeline” (source, qualify, write, follow up, measure) shows “Not offered by NeoRank” at each step.
- An engine's trust in an outlet: no defensible measure exists; NeoRank shows no score in its place.