Tonecast for PR and communications teams
For a PR or communications team, Tonecast shows how AI assistants describe your organisation, product or spokesperson, which claims they repeat, and which articles, threads and videos carry each story. When a claim spreads in the sources before it reaches the answers, and the pattern is statistically reliable, Tonecast flags it early.
Tonecast is a SaaS platform that shows what AI assistants (ChatGPT, Claude, Gemini and Perplexity) say about a brand, product, person or topic, and which web pages shaped those answers. This page is for the people who manage reputation and narrative: press offices, corporate communications, public affairs and the agencies that support them.
Which questions does Tonecast answer for PR and communications?
| Question | Where the answer is |
|---|---|
| How do AI assistants describe us, and in what tone? | Tone in AI on the Overview; the full text of every answer in AI Answers |
| Which claims do they repeat about us, and are they true? | Claims view, against your fact sheet |
| Which articles, threads and videos carry a story? | What shaped it: pages with their tone, claims and influence weight per engine |
| Is a narrative spreading in the sources before it reaches AI answers? | Early signals: claim prevalence and the Formative Sentiment Index, per engine |
| How did a launch or an announcement land with AI? | Launch Mode: before and after comparison (Starter plan and above, rolling out) |
| What should we respond to first? | Actions: negative high-weight sources and wrong claims |
Why does tone in AI matter for communications?
AI assistants rarely turn openly hostile. A more common problem is quieter: a brand drops out of recommendations, is described with a hedge ("some users report…"), or is paired with an old controversy. Tone in AI measures the tone towards your subject specifically, not the tone of the whole answer, and every score opens the sentences behind it.
How are claims tracked?
A claim is a factual statement: "the company was fined", "the product was discontinued", "prices rose in 2026". Tonecast extracts claims from AI answers and from the source pages, and measures claim prevalence: how much of the weighted source graph supports each claim over time. A claim that grows in the sources and is not yet in the answers is exactly the kind of signal a press office wants to see early.
Can Tonecast warn us before AI answers change?
Sometimes, and it tells you when it cannot. For each engine, Tonecast tests whether movements in the sources came before movements in the answers. The state is observing until there is enough data and the pattern is statistically reliable; only an active signal becomes an early warning, with the estimated lead time in days. See the methodology for the exact rule.
Can we track a spokesperson or an issue?
Yes. Campaigns can be about a public person or a general topic as well as a brand or product. Tonecast proposes the questions people ask about each ("who is…", "what does … think about…", "what is happening with…") and disambiguates namesakes with aliases and exclusions you confirm.
What does Tonecast not do?
It does not monitor closed social platforms, and it does not write responses or corrective content. It shows what AI says, which pages are behind it, and where a response would matter most.