How it works

How Radar finds who AI cites, and what to publish next.

Six parts, one loop: prompts, engines, citations, share of voice, fixes, and a weekly cadence.

1. Prompts

The prompt set is the buyer's question list.

Each prompt is a question a buyer might ask an AI assistant before buying. We write them from your product, your category, and your competitors.

best HIPAA compliant form builderform builder with audit logsembeddable forms for clinicsform tool with e-signatures

2. Engines

Each prompt runs on four engines.

Every prompt goes to ChatGPT, Gemini, Claude, and Perplexity. We record each answer as it is returned, without changing it.

ChatGPT
Gemini
Claude
Perplexity

3. Citations

A citation is a named brand in the answer.

A green check means the engine named you. A gray dash means it did not. Red is reserved for alerts, such as a competitor overtaking you.

Citation table

Sample data

Sample citation results for four prompts across four AI engines
Prompt	ChatGPT	Gemini	Claude	Perplexity	Top competitor
best HIPAA compliant form builder					Competitor A
form builder with audit logs					Competitor B
embeddable forms for clinics					Competitor A
form tool with e-signatures					Competitor C

Green check: cited. Gray dash: not cited.

4. Share of voice

It is the share of engine runs that name your brand, shown next to your competitors.

Sample data

40%

19 of 48 engine runs cite you.

You40%

Competitor A52%

Competitor B27%

5. Fixes

Every gap comes with a page to publish.

Each fix names the page, the priority, and the prompt it targets.

Prioritized fix list

Sample data

P1/hipaa-form-builder (new page)
Own the intent: publish a dedicated page matching 'HIPAA compliant form builder'.
Targets prompt: HIPAA compliant form builder
P2/pricing (update)
Add an audit log section that answers the question directly.
Targets prompt: form builder with audit logs
P3/clinics (new guide)
Publish a clinic setup guide with a short answer at the top.
Targets prompt: embeddable forms for clinics

How Radar finds who AI cites, and what to publish next.

The prompt set is the buyer's question list.

Each prompt runs on four engines.

A citation is a named brand in the answer.

Citation table

Every gap comes with a page to publish.

Prioritized fix list

Paid tiers rerun the loop every week.

See your own citation table.

How Radar finds who AI cites, and what to publish next.

The prompt set is the buyer's question list.

Each prompt runs on four engines.

A citation is a named brand in the answer.

Citation table

Share of voice is your percentage of cited runs.

Share of voice

Every gap comes with a page to publish.

Prioritized fix list

Paid tiers rerun the loop every week.

See your own citation table.