CitedOS measures you against a prompt set (the buyer questions your category gets asked) and a competitor set. Setup pre-fills both so you start from a real baseline, not a blank page.
What setup pre-fills
When you create a workspace, CitedOS enriches your brand from its domain — name, aliases, industry, location and language — then suggests a starter prompt set and discovers likely competitors in your category. You keep, edit or delete each one. Nothing is locked.
AI-suggested prompts
Buyer-intent questions generated from your category, e.g. "best CI/CD tool for startups" or "Segment alternatives". These are suggested prompts; you can add your own user-defined prompts alongside them.
Discovered competitors
The rivals most likely to be named for the same questions. Each competitor is measured on the same prompts so every comparison is apples-to-apples.
How many prompts?
- Aim for at least 10 prompts — enough breadth to read a real visibility picture across your category, not one lucky query.
- Write prompts the way a buyer would type them — comparison, alternatives, "best X for Y", problem-first questions. Brand-name searches don't reflect discovery.
- Each prompt is run across all three engines, multiple times — so prompt count, not engines or runs, is what your plan's allowance counts. See Billing & plans.
Choosing competitors
Pick the brands you actually lose deals to and the ones AI tends to name for your category — they're often not the same list. CitedOS shows where each gets cited that you don't, which is the raw material for the source-gap report.