Skip to main content
Rankahead introduces a set of concepts that may be new if you are coming from traditional SEO tools. This page explains the core terminology so you can move confidently through the platform. Each concept links to the relevant section of the documentation where you can read more.

Domains and tracking

Domain

A domain is a website or brand that you track inside Rankahead. It is the central unit of organization — prompts, competitors, and analytics are all scoped to a domain.When you create a domain you provide three things:
  • Display name — a human-readable label used throughout the interface.
  • Website URL — the canonical URL of the brand’s site, used to identify citations in AI responses.
  • Brand name — the exact string Rankahead looks for inside LLM-generated answers to detect a mention.
Each plan has a domain limit: Free (1), Solopreneur (3), Agency (unlimited). You manage domains from Settings → Domains.
A competitor is another brand that you add to a domain for benchmarking purposes. When Rankahead analyzes AI responses, it looks for both your brand and all configured competitors. This data feeds into share-of-voice calculations and answer gap analysis.You can add competitors during onboarding or at any time from Settings → Domains → [your domain] → Competitors. Each competitor requires a brand name and website URL.

Prompts and runs

Prompt

A prompt is a query that Rankahead sends to one or more LLM providers to check whether your brand appears in the answer. Prompts represent the questions your potential customers might type into ChatGPT, Perplexity, Google AI Overviews, or Anthropic.A prompt has four key fields:
FieldPurpose
NameInternal label for identifying the prompt in the dashboard
Query textThe exact question submitted to the LLMs
Brand nameThe string to look for in responses (inherited from the domain)
ScheduleA cron expression or preset interval that controls when the prompt runs automatically
Prompts are managed from Dashboard → Prompts. You can also trigger a prompt manually at any time with the Run now action.
Write prompts in the same language and phrasing your customers use. Questions like “What is the best tool for X?” or “How do I solve Y?” tend to produce more realistic visibility data than abstract or keyword-heavy queries.
A prompt run is a single execution of a prompt. When a run starts, Rankahead sends the query text to every configured LLM provider simultaneously — OpenAI, Anthropic, Google, and Perplexity — and records each response. After collecting all responses, it extracts mentions, citations, and competitor appearances.Runs can be triggered in two ways:
  • Scheduled — automatically on the cron schedule you configure for the prompt.
  • Manual — on demand by clicking Run now in the interface or calling the API.
Both types count equally toward your plan’s monthly prompt quota. You can inspect the full raw response from each LLM provider by clicking into any individual run result.

Visibility metrics

Visibility score

The visibility score is a 0–100 number that summarizes how consistently your brand appears in AI responses across all prompt runs for a domain. A score of 100 means your brand was mentioned in every response from every LLM for every prompt. A score of 0 means your brand did not appear in any response.The score is calculated from a weighted combination of mention rate and response position (brands mentioned earlier in a response score higher). It is designed to give you a single number you can track over time and compare across domains or competitors.
Visibility scores are provider-specific as well as aggregate. You can filter the dashboard to see your score broken down by ChatGPT, Perplexity, Google AI Overviews, or Anthropic separately.
The mention rate is the percentage of prompt runs in which your brand was mentioned at least once by a given LLM provider. For example, if you ran a prompt 10 times and ChatGPT mentioned your brand in 7 of those runs, your mention rate for that prompt on ChatGPT is 70%.Mention rate gives a raw measure of how reliably a specific LLM includes your brand when answering a particular type of question.
Share of voice measures your brand’s proportion of total brand mentions in AI responses, compared to all tracked competitors. If your brand and three competitors were each mentioned in AI responses and your brand accounted for 40 of the 100 total mentions across all runs, your share of voice is 40%.Share of voice helps you understand your competitive position in the AI channel, not just your absolute visibility. You can view it in aggregate or broken down by LLM provider and individual prompt.

Gap analysis

Answer gap

An answer gap is a query or topic where one or more competitors appear in AI-generated answers but your brand does not. Answer gaps represent missed opportunities — questions that real users are asking AI tools where your brand is invisible.Rankahead identifies answer gaps automatically by comparing mention data across your brand and all configured competitors for every prompt. You can review them in Dashboard → Gap Analysis, sorted by competitive exposure and potential impact.Closing an answer gap typically means creating or improving content so that LLMs learn to associate your brand with that topic. GEO tasks (see below) help you track the actions required to close each gap.
A citation is a domain URL that an LLM references or links in its response — either as a source, a recommendation, or a suggested next step. Citations matter because LLMs learn from the content they index; being cited frequently signals topical authority.Rankahead extracts citations from every prompt run response and tracks which domains are cited most often in answers related to your prompts. You can view citation data in Dashboard → Citations. Building citations — getting your domain referenced in high-quality content that LLMs surface — is one of the primary GEO optimization strategies.

Optimization

GEO task

A GEO task is a specific optimization action tracked on a kanban board in the Content → GEO Tasks section. Tasks represent the concrete steps you take to improve your AI visibility — such as writing a blog post targeting an answer gap, adding schema markup, or earning citations from authoritative sources.Each GEO task has a category that describes the type of optimization:
CategoryDescription
CONTENT_GAPCreate or improve content to target a query where competitors appear but you do not
SCHEMA_MARKUPAdd structured data markup to help LLMs extract and surface information about your brand
CITATION_BUILDINGGet your domain referenced in content that LLMs are likely to surface in relevant answers
FORUM_PRESENCEParticipate in forums, communities, and Q&A sites where LLMs draw answers from
Tasks move through kanban columns (for example, To do → In progress → Done) so you can track the status of your GEO program at a glance.

Integrations

Rankahead can connect to your Google Search Console account to import traditional search performance data alongside your AI visibility metrics. This lets you identify striking-distance keywords — queries where your pages rank on page 2 or 3 of Google — and correlate traditional search rankings with AI visibility trends.You connect Google Search Console from Settings → Integrations. Once connected, the data appears in Dashboard → Search Performance.
A CMS integration connects Rankahead to your content management system so you can publish AI-generated blog posts directly from the Content section without copying and pasting. Rankahead pushes the generated post — including title, body, and metadata — to your CMS with one click.You configure CMS integrations from Settings → Integrations. See CMS integrations for setup instructions.
An MCP token is a Bearer authentication token that you use to connect an AI agent to Rankahead’s Model Context Protocol (MCP) server. The MCP server exposes Rankahead’s data and actions as structured tools that any MCP-compatible AI agent — such as Claude, Cursor, or a custom agent — can call programmatically.The MCP server endpoint is POST https://app.rankahead.ai/api/mcp. You authenticate by including your token in the Authorization header:
Authorization: Bearer YOUR_MCP_TOKEN
You generate and manage MCP tokens from Settings → API Keys. See MCP integration for the full list of available tools and example workflows.
Treat your MCP token like a password. Anyone with the token can read your visibility data and trigger prompt runs on your account. Rotate it immediately from Settings → API Keys if you suspect it has been exposed.

Next steps

Quickstart

Follow the step-by-step guide to create your first domain, prompt, and run.

Dashboard overview

See how visibility scores, mention rate, and share of voice appear in the analytics dashboard.

GEO tasks

Learn how to create, categorize, and manage GEO tasks to track your optimization work.

MCP integration

Connect an AI agent to Rankahead for automated visibility monitoring and reporting.