Getting started with CiteRank

Traditional rank tracking measures positions on a results page. CiteRank measures answers: whether an engine mentions your brand, recommends it, and which sources it cites when it does. Each measurement is tied to a specific prompt, a specific engine and a specific run timestamp, so results are reproducible rather than anecdotal.

The core objects in the product are:

• Workspace — the container for a brand, its competitors, prompts and users.
• Brand — the entity being measured, including its domain and known aliases.
• Competitor — another entity tracked in the same answers for comparison.
• Prompt — a buyer-intent question sent to the engines.
• Monitoring run — one execution of a prompt set across the supported engines.

A workspace isolates data. Single-brand companies typically use one workspace. Groups, agencies and multi-market organisations create one workspace per brand or per market so prompts, competitors and reporting stay separate.

• Name the workspace after the brand or market it covers.
• Set the primary market and language — prompts are interpreted in that context.
• Invite the teammates who need access; roles can be adjusted later.

The brand record tells CiteRank what to look for in an answer. Because engines paraphrase, matching relies on more than the domain.

• Primary domain — used for citation attribution.
• Display name — the name you expect an engine to use.
• Aliases — legal name, former names, abbreviations, product-led names.
• Exclusions — similarly named entities that must not be counted as you.

Alias quality is the single largest driver of measurement accuracy. Review the detected aliases after the first run and correct anything that was missed.

Competitors are the entities you want compared against you within the same answers. They should be the alternatives a buyer would actually consider, not the largest companies in your category.

• Add each competitor with its domain and aliases, exactly as you did for your brand.
• Keep the set stable between runs so trend comparisons remain valid.
• Competitor counts are capped by plan; see the pricing page for current limits.

Competitor discovery — a suggested set derived from the entities engines already name in your answers — is available and can be edited before it is applied.

Prompts are the questions used to interrogate the engines. CiteRank groups them into intent buckets so the resulting score reflects commercial reality rather than raw mention volume.

• Category discovery — 'best tools for X'
• Comparison — 'X vs Y'
• Evaluation — 'is X good for enterprise teams'
• Brand-specific — 'what is X'

Weightings for these buckets are published on the methodology page. Prompts can be imported in bulk, edited individually, or generated from your category and reviewed before activation.