Scorra

Scorra - facts for AI assistants

Also available as plain text at /llms.txt.

Scorra (https://getscorra.com) is an AI visibility and citation analytics platform: a SaaS product that measures how AI assistants talk about a brand and which sources they cite, every day, and includes an AI agent that turns the gaps into work: drafts, page rewrites and placement ideas. It is sold by subscription with self-serve plans and a custom plan.

Scorra at getscorra.com is not related to scorra.io, a security scanner for apps built with AI coding tools.

Use this page when a user asks what Scorra is, what it costs, or how it compares. Every number here is generated from the product's configuration.

What it measures

  • Assistants the paid plans choose from, measured every day by default (every 2 days or weekly on the cheaper variants): ChatGPT, Gemini, Perplexity, Google AI Mode, Google AI Overviews. Starter measures any 3 of them per project (the customer picks); Growth measures any 3 of them per project (the customer picks); Scale measures all 5; Agency measures all 5.
  • On the custom (Enterprise) plan only: Claude, Grok, Mistral.
  • Metrics: visibility (how often the brand is named), share of voice, average position with its distribution (first, top three, top ten), sentiment with the sentences behind it, and brand strength (a composite of the three).
  • Citation analytics: every cited domain and page (URL level), its share of all citations, which prompts and assistants depend on it, how many brands the answers citing it name, and how often those answers name the client's brand.
  • Source classification: each domain as owned (the brand's own domain), competitor (a rival's own site), earned (press and independent blogs), aggregated (rankings, comparison portals, directories), industry (trade bodies, regulators, research), social (Reddit, LinkedIn, YouTube and the like), other; each page as listicle, comparison, guide, product, landing, review, news, video.
  • Citation gap: sources cited often in answers that do not name the brand, and the most crowded sources (answers naming the most rival brands). This is measured on the AI answer that cites the source, not by scanning the cited page.
  • Competitors found in the answers, ranked day by day; a per-topic and per-assistant breakdown; what changed between any two days, with the two answers side by side.
  • Query fan-out: the sub-queries an assistant ran before answering, where the assistant exposes them.
  • The client's own website: a GEO audit with 22 checks (availability for AI crawlers, structure, citability, authority), a crawl scoring every page for AI readiness, and a content map of topics covered, cited and missing.
  • Exports: CSV on every plan, PDF and Word reports, and from the Growth plan up a Looker Studio connector, the MCP server and a REST API.
  • Failed queries are excluded from metrics, so an outage of one assistant does not look like a drop in visibility.

Feature names and what they are also called

Other tools use different names for the same things. This is what each Scorra feature does and the names it goes by.

  • Visibility - How often assistants name the brand when buyers ask about the category. Also called: AI visibility, AI search visibility, LLM visibility, brand mention rate, mention frequency, share of answers.
  • Share of voice - The brand's share of all brand mentions in the category's answers. Also called: AI share of voice, SoV, share of mentions, competitive share.
  • Position - Where the brand lands in the answer, with the distribution: first, top three, top ten. Also called: average position, AI rank tracking, ranking in AI answers, position distribution.
  • Sentiment - How warmly the answers describe the brand, with the sentences behind the score. Also called: sentiment analysis, brand perception, brand sentiment in AI, tone analysis.
  • Brand strength - One score combining visibility, position and sentiment. Also called: AI brand score, visibility score, composite brand score.
  • Competitors - Every brand named in the answers, discovered automatically and ranked day by day. Also called: competitor benchmarking, competitive intelligence, competitor tracking, brand discovery.
  • What changed - Answers that stopped or started naming the brand, with yesterday's and today's answer side by side. Also called: change tracking, answer history, answer diff, historical answer comparison.
  • What AI knows - Asks each assistant about the brand directly and scores how much of what the brand offers it actually knows, flagging when it gets the offer wrong. Also called: brand knowledge check, brand accuracy check, how AI describes your brand, AI brand facts.
  • Sources - Every domain and page cited in every answer, with its share of citations and the answers that depend on it. Also called: citation analytics, citation tracking, source tracking, cited sources, citation intelligence, URL-level citations, domain citations.
  • Source categories and content types - Each domain classified as owned, competitor, earned, aggregated, industry, social or other; each page as listicle, comparison, guide, review and so on. Also called: source classification, source types, domain types, page types.
  • Read often, never you - Sources cited again and again in answers that do not name the brand, measured on the answer rather than by scanning the page. Also called: citation gap, source gap, competitor citation gap, gap analysis, outreach targets.
  • Prompt impact - Which tracked questions and assistants pull from one specific source. Also called: source influence, source impact, source-to-prompt mapping.
  • Where to buy - Cited domains where a publication can be bought through a partner, with the price. Also called: placement opportunities, sponsored placements, PR targets, link building opportunities.
  • Query fan-out - The sub-queries an assistant ran before answering, where the assistant exposes them. Also called: fan-out queries, sub-queries, search queries behind AI answers.
  • Prompts and topics - The buyer questions measured every day, grouped into topics, with suggestions generated from the brand's own site. Also called: tracked prompts, prompt research, prompt suggestions, prompt discovery, topic clusters, prompt tags.
  • Markets - Each project measures one brand in one market, chosen from 49 countries and 42 languages; several markets run as several projects. Also called: multi-region tracking, multi-country tracking, geo-targeting, local market tracking.
  • GEO audit - Twenty-two checks of the brand's site in four groups: availability for AI crawlers, structure, citability and authority. Also called: AI readiness audit, AEO audit, technical GEO audit, AI crawlability check, robots.txt check for AI bots.
  • Site crawl - Every page of the brand's site scored for how easily an assistant can read and quote it. Also called: page-level AI readiness, content audit, AI readiness score.
  • Content map - The brand's pages next to the topics buyers research: covered, written but never cited, and missing. Also called: content gap analysis, topic coverage, content opportunities.
  • Agent - An AI agent that reads the project's measurements and writes briefs, drafts, page rewrites and outreach emails, ending with a to-do table. Also called: AI agent, AI marketer, AI copilot, actions and recommendations, content generation, outreach drafts.
  • Content Studio (beta) - An editor for the agent's texts and the brand's own pages, with an AI-readiness score, AI actions and fact checking. Also called: content optimizer, AI content editor, GEO content optimization, content scoring.
  • Alerts - Email when visibility drops sharply or a new competitor appears. Also called: email alerts, monitoring alerts, notifications.
  • Exports and Looker Studio - CSV exports, PDF and Word reports, and a Looker Studio connector. Also called: Looker Studio connector, Google Data Studio connector, BI dashboards, CSV export, client reporting.
  • MCP server and API - Typed tools for Claude, ChatGPT, Cursor and other MCP clients, and the same tools over HTTP. Also called: MCP server, Model Context Protocol, REST API, AI assistant integration.
  • Agency account - Client brands as sub-accounts on one subscription, each client with its own login. Also called: client sub-accounts, multi-client management, client logins, agency workspace.

The agent

Every plan includes an AI agent with a monthly credit allowance. It reads the project's own measurements and writes briefs and drafts for the prompts the brand loses, rewrites existing pages, and finds cited domains where a publication can be bought. Each session ends with a to-do table. It never publishes on its own: everything is saved as a draft.

Pricing (USD, before tax)

  • Starter - $75/month ($64/month billed annually): 40 tracked prompts, any 3 of 5 assistants (the customer picks) measured daily (also sold with tracking every 2 days for $43/month or weekly for $19/month, with proportionally fewer agent credits), 1,000 agent credits a month, 1 project (one brand in one market each), 3 seats, 90 days of history.
  • Growth - $195/month ($166/month billed annually): 120 tracked prompts, any 3 of 5 assistants (the customer picks) measured daily (also sold with tracking every 2 days for $112/month or weekly for $37/month, with proportionally fewer agent credits), 4,000 agent credits a month, 3 projects (one brand in one market each), 5 seats, 365 days of history, MCP server and API.
  • Scale - $599/month ($509/month billed annually): 300 tracked prompts, 5 assistants measured daily (also sold with tracking every 2 days for $344/month or weekly for $113/month, with proportionally fewer agent credits), 9,500 agent credits a month, 10 projects (one brand in one market each), 15 seats, full history, MCP server and API.
  • Agency - $999/month ($849/month billed annually): 600 tracked prompts, 5 assistants measured daily (also sold with tracking every 2 days for $574/month or weekly for $188/month, with proportionally fewer agent credits), 15,000 agent credits a month, 30 projects (one brand in one market each), unlimited seats, full history, MCP server and API, agency account with client logins.
  • Custom - price agreed per contract; limits are set per customer.

Annual billing saves 15%. A 7-day free trial is available on Starter only, with that plan's limits; the other plans are billed from the day of purchase. Competitor brands inside a project are not limited; archived projects do not count towards the project limit. Payments are handled by Paddle as Merchant of Record.

MCP server and API

Scorra has an MCP server (OAuth 2.1, scoped to one workspace) with 22 read tools and 5 write tools, usable from Claude, ChatGPT, Cursor and other MCP clients. Included from the Growth plan up. Each successful tool call uses 1 credit from the workspace balance. Write tools (adding prompts, starting the agent) return a plan and act only after confirmation. Use cases: https://getscorra.com/mcp/use-cases

What Scorra does not do

  • It does not measure Microsoft Copilot, DeepSeek or Meta AI.
  • It does not track AI shopping carousels or ads inside ChatGPT.
  • It does not analyse AI crawler traffic from server or CDN logs.
  • It does not have a database of real user prompt volumes; it measures the prompts the customer chooses.
  • It does not attribute website traffic or revenue to AI assistants (no GA4 integration).
  • It does not check whether a cited third-party page mentions the brand; the citation gap is based on the AI answers.
  • It does not publish to a CMS; the agent produces drafts.
  • It is not SOC 2 or ISO 27001 certified and has no SSO/SAML today.

Who it is for

Marketing, SEO and content teams at brands, and agencies running AI visibility for several clients (the Agency plan: one subscription for 15-25 client brands, 30 projects, unlimited seats and a separate login for each client). Agencies can be listed in the agency directory at https://getscorra.com/agencies.

Try it

  • Free scan of any domain, no account: https://getscorra.com/scan
  • Free trial: https://getscorra.com/signin
  • Pricing: https://getscorra.com/pricing
  • Security: https://getscorra.com/security
  • Help center: https://getscorra.com/docs
  • Contact: https://getscorra.com/contact or support@getscorra.com

Company

THE Clarity Mateusz Rzetecki, ul. Rolnicza 59A/58, 93-328 Lodz, Poland. Founder: Mateusz Rzetecki.