Meet Echo: Know What AI Says About You

People now ask ChatGPT, Gemini and Claude before they ask you. A customer choosing a product, a journalist checking a story, an investor weighing a company and a voter making sense of an issue all get the same thing back: not a page of links to judge for themselves, but one confident answer. That answer names some brands and leaves out others, takes a view, and quotes sources the reader rarely checks.

For most organizations that answer is a blind spot. It is not on their website, it does not show up in their media monitoring, and it changes when the models are updated. Two people asking the same question of two assistants can get different stories.

Echo is Vinesight's product for closing that blind spot. This post explains what it does, who it is for, and what sets it apart.

Key takeaways

• Echo shows what AI says about you. It puts your questions to several dozen AI models every day and records who they recommend, what they say and where they get it.
• It is built for every team that owns a story: communications, brand and reputation, marketing and SEO, public affairs, executive offices and investor relations, and risk.
• It knows the questions before they are asked. Echo uses Vinesight's social media intelligence to predict the questions people are about to put to AI, and tracks them from the moment they emerge.
• It measures like a panel, not a spot check. The same questions every day, as different audiences and in different languages, with every number traceable to the answer behind it.

What Echo does

Echo works like a daily survey of AI. You choose the questions, Echo asks them, and it reads every answer.

It asks three kinds of question. Questions about you: whether you are trustworthy, how you compare with a named competitor, what you are known for on pricing, safety or sustainability. Unbranded questions that never name you, such as "which bank should I trust with my savings?", to see whether AI brings you up unprompted. And contested questions, where you define the positions that matter and Echo sorts every answer into them.

It asks every major model, every day. Several dozen models from the leading AI companies in the United States, Europe and China: the assistants people use, such as ChatGPT, Gemini, Claude, Perplexity, Grok and DeepSeek, and the models other products are built on. Questions go out with and without web search, because a model answering from memory and a model reading today's news can tell very different stories.

It reads every answer for four things:

• Visibility. Are you named, how often, and in what position? When someone asks AI to choose, are you the recommendation, one of the options, or missing?
• Stance. Is the answer supportive, neutral or critical? Echo scores each answer and rolls them up into a net sentiment from −100 to +100, by model, by question and over time.
• Talking points. The praise, criticism and recurring themes each assistant repeats, so you know which story each one tells.
• Sources. Every link and publication the models cite, grouped into news, government, reviews, social media and your own website, so you know which pages shape the answer.

The results sit in a live dashboard, updated daily, alongside reports written for the team that has to act on them.

Who Echo is for

Anyone whose job depends on how an organization, a person or an issue is described now has a new audience to watch. Echo is built for the teams that own the story.

Communications and PR

When a reporter, a customer or a critic asks an AI assistant about you, the answer is a briefing you did not write. Echo shows what that briefing says, which criticisms it raises and which sources it leans on, so you can correct the record at the source rather than one conversation at a time. When a story breaks, it shows how quickly AI picks it up and how each model frames it.

Brand and reputation

Reputation used to be measured in surveys, coverage and search results. Echo adds the answer people now trust most. It tracks how AI describes your brand, your products and your leaders against your competitors, which old controversies keep resurfacing, and whether the story is improving week by week.

Marketing, SEO and digital

Search is becoming a conversation, and the conversation ends in a recommendation. Generative engine optimization, or GEO, is the work of earning that recommendation, and it starts with measurement. Echo shows where you rank when people ask AI to choose a product or a provider, which competitors win the top spot, and which pages the models trust. That tells you what to fix, what to publish and where to focus, and gives you a daily trend to show whether it is working.

Public affairs and policy

On contested issues, AI does not only inform: it frames. Echo lets you define the positions that matter on an issue and sorts every answer, from every model, into them. You can see which assistants present one side, which present both, and how the framing shifts with the language of the question or the country the model was built in.

Executive offices and investor relations

Investors, analysts and prospective hires ask AI about companies and the people who run them. Echo tracks how assistants describe your company, its strategy and its executives, and flags when a criticism or an outdated fact starts to stick.

Risk and intelligence

AI answers can be shaped by whoever fills the web with content the models read. Echo's source tracking shows when fringe, partisan or state-backed outlets start to appear behind the answers about you or your issues, so risk and intelligence teams can see when an influence effort has reached AI, and which models it has reached.

What makes Echo different

1. It knows the questions before they are asked

Most tools that measure AI answers check a list of prompts someone wrote in advance. That list is only as good as the person who wrote it, and it is out of date the moment something happens.

Echo starts from what people are actually saying. Vinesight's narrative intelligence watches the conversation across dozens of social media platforms, mainstream and fringe, in more than 50 languages. When a story about you starts to build there, whether it is a viral video, a forum thread or a boycott call, Echo infers the questions people are about to put to AI: "Is this brand safe?", "Why are people boycotting this company?", "Is this claim true?"

Those questions are tracked from the moment they emerge, across every model. You see how AI answers them while the answer is still forming, and before most of your audience has asked. That is the difference between finding out what AI said about a crisis afterwards and shaping what it says during one.

It works in both directions. What Echo finds in AI answers, such as a source that keeps appearing or a criticism that keeps resurfacing, tells the narrative team what to look for on social media next.

2. It measures like a panel, not a spot check

A single screenshot of a chatbot answer proves very little: ask again and the answer may change. Echo treats AI like a panel survey. Once a dashboard starts, its questions are locked and never reworded, and each one is asked every day. A change in the answer is a change in AI, not in the question.

Results are weighted so that no single question or model dominates, and every score comes with the answers behind it.

3. It asks as your audiences, in their languages

AI assistants increasingly remember who they are talking to, and that changes the answer. Echo can ask the same questions as different kinds of users, such as an investor, a parent or a first-time buyer, by giving the model a short profile the way consumer chatbots use their memory of a user. You see how much the recommendation depends on who is asking.

It can also ask in other languages, to compare what AI tells people in each of your markets. On many issues the answer in one language is not the answer in another.

4. It follows the sources down to the passage

Knowing that a model cited a website is useful. Knowing which sentence that website supports is what lets you act. Echo records each link with the passage that cites it and whether that passage is favourable or critical, and checks whether the link still works. You learn which pages feed the criticism, which outdated articles still frame the answer, and whether your own website is saying what you need it to say. Brand sites are often among the most cited sources of all.

5. Every number traces back to an answer

Echo's scores are read by an AI classifier, and it is built to be checked. The classifier sees only the question and the answer, never the user profile the question was asked as, so a persona cannot bias the score. Every percentage and every sentiment score links to the answers and quotes behind it, so a finding can be verified before it reaches a board pack.

6. It comes with the team that reads it

Echo is part of Vinesight's narrative intelligence platform, built by the same analysts who track narratives and influence operations for Fortune 500 companies and governments. Alongside the live dashboard, clients get regular reports with clear recommendations: which criticisms to answer, which sources to fix or feed, and where AI visibility can be won.

From insight to action

What a team does with Echo depends on what it finds:

• A criticism keeps surfacing. Find the sources behind it, correct or answer them, and watch whether the models follow.
• A competitor wins the recommendation. See which pages the models trust for that question, and publish the content that earns the answer.
• A story is building on social media. Check how AI already answers the questions it will raise, and prepare before the questions arrive.
• Your own site is a top source. Make sure the pages AI quotes about you are current, clear and saying the right thing.

In each case Echo shows whether it worked, because the same questions are asked again the next day.

See what AI says about you

AI assistants already have an opinion of your brand, your leaders and your issues. Echo tells you what it is, where it comes from, and what people will ask next. Explore Echo, or request a demo below and we will show you what several dozen AI models say about you.