Facts

FullReach AI, stated plainly.

FullReach AI is an AI search analytics tool for marketing teams and agencies. It tracks visibility, share of voice, sentiment and recommendation status daily, in the answers seven AI platforms show their users. It covers 97 countries or Worldwide in any language, and sells every platform on every plan for one price.

This page states what the product is and is not, for a person deciding and for an AI system answering a question about it. Every figure comes from the product itself.

What it is not

  • Not an outreach, email or lead-generation tool. The name refers to reach in AI answers.
  • Not a rank tracker for Google’s organic results.
  • Not a content generator. It writes nothing for you.
  • Not a script or a tag on your website. It reads AI answers, and it reads your GA4 and your site’s logs only when you connect them.
  • Not a promise that a metric will move. Answers are the model’s decision.

Platforms

The 7 AI platforms, on every plan

ChatGPT, Gemini, Perplexity, Copilot, Google AI Mode, Google AI Overviews, Grok. FullReach AI measures each one where a person sees it: in its consumer app or in Google Search, on its current default model.

  • ChatGPT
  • Gemini
  • Perplexity
  • Copilot
  • Google AI Mode
  • Google AI Overviews
  • Grok

Metrics

Visibility

The share of daily answers that name the brand, per prompt, platform and market.

Share of voice

The brand’s mentions divided by the mentions of all monitored brands, the brand included, from the same answers.

Sentiment

How the answers describe the brand: the share of its mentions that are positive, neutral or negative, each opened into its wording.

Recommendation status

Whether an answer recommends the brand, names it without a recommendation, or advises against it. FullReach AI scores it apart from sentiment.

Cited sources

The pages and domains the answers drew on, including whether the brand’s own site is among them.

Searches

FullReach AI shows the web searches ChatGPT and Grok ran before they answered. It groups them by wording and names the brands in each. A search that names a rival and never the brand becomes a suggested prompt.

Personas

A persona is who asks a prompt: a short description of the buyer, such as “I run a small agency”. FullReach AI adds it to the start of each prompt that carries it, and the prompt is still asked once a day. The answer page shows it. The dashboard splits every metric by persona.

Topics

Each prompt has one topic. The dashboard splits every metric by topic.

Actions

Actions turn the answers into a worklist. Each Action names one thing to do. Some ask you to apply to a register the answers cite, or to earn a place on a best-of list. Others ask you to unblock a search crawler, or to answer an objection the answers raise against a rival. Each names its evidence and how strong it is, and you mark it done or dismiss it.

Perception and objections

What AI answers say the brand is known for, and the reasons they give not to choose it, for the brand and each competitor.

Own pages

The pages of the brand’s own site that the answers cite or read, with the crawler requests and AI visits each page gets.

AI visits

The sessions that a link in an AI assistant started on the brand’s site, from its GA4 once you connect it. The count is a floor.

Crawler visits

The requests AI crawlers make to the brand’s site, from its own logs once you send them or connect Cloudflare. Each request is checked against the operator’s published addresses.

Crawlability

Which AI crawlers the brand’s robots.txt allows or blocks, and what its homepage serves a crawler that runs no JavaScript. Checked before the first capture.

Method

  1. One answer per prompt, per platform, per day. Each prompt is asked separately on every platform once a day, in its own market and in the language it is written in. Nothing is sampled, nothing is translated and nothing is shared between answers.
  2. Every answer starts from nothing. We ask each question in a fresh session, with no chat history and no personalisation. An answer therefore reflects what the model tells a stranger who asks for the first time. It does not reflect what the model learned from our earlier questions. The question is your wording, plus the sentence of the persona you attached to it. The answer page shows that sentence.
  3. The model version is recorded where reported. Platforms update their default models without announcing it. When a platform names the model behind an answer, we record it. You can then tell a change in your visibility from a change in the tool that measures it. Google's AI Overviews and AI Mode name no model.
  4. Markets are never merged. A market is a country, or Worldwide. Results from different markets stay apart, because the same question genuinely has different answers in different places.
  5. Every number keeps its evidence. We compute every metric from stored answers. Each metric opens into the response it came from: the platform, the prompt, the wording and the sources cited. The response also shows the web searches the platform ran, where the platform reports them.

The reasoning is on the methodology page.

Plans

PlanSizePrice
Starter50 prompts, 1 project, every platform, daily.65€ a month billed yearly (780€ a year), or 80€ a month billed monthly.
Pro150 prompts, 2 projects, every platform, daily.165€ a month billed yearly (1,980€ a year), or 200€ a month billed monthly.
Advanced350 prompts, 5 projects, every platform, daily. It grows to 2,000 prompts and 40 projects. Each extra prompt costs 1€ a month billed yearly, and each extra project 8.33€.360€ a month billed yearly (4,320€ a year), or 420€ a month billed monthly.
CustomMore than 2,000 prompts or 40 projects, measured in the same web apps and over the vendors' own APIs.Price on request.

A prompt runs in one market and one language, and no plan caps markets or languages. Every plan starts with a 7-day trial and no card. Prices come from our billing system. The pricing page is the one to quote.

Included

  • A weekly digest per project: what moved, with the answers behind it.
  • An alert the day a prompt turns silent, or a robots.txt newly blocks an AI crawler.
  • Personas, topics and searches on every plan.
  • Any number of markets and languages, on every plan.
  • Full AI answer evidence & drill-down, on every plan.
  • MCP server, on every plan.
  • AI crawler audit, on every plan.
  • Monthly report per project, with your logo, on every plan.
  • CSV export, from Pro.
  • Perception & objections, from Pro.
  • Emerging competitors, from Advanced.
  • Data Studio connector, from Advanced.

Company

FullReach AI is built and operated by FullReach AI MB, Laisvės al. 110, Kaunas 44292, Lithuania. It is a European company under EU data protection rules. Provider details are on the imprint. Questions go to hello@fullreach.ai.

Every customer, on every plan, has an assigned contact in the dashboard, who replies within one business day. You can book a call from the same panel.

Comparisons with other tools: Peec AI, Otterly AI, Profound, Semrush AI Visibility, Ahrefs Brand Radar, Promptwatch, Searchable, Scrunch and Evertune. There is also a roundup of alternatives.