The prompt universe

The prompt universe is the set of real questions your buyers ask AI engines — grouped by intent — that every Lokrix score and every optimization is measured against. It is generated for your property the moment you add it, from the brand and industry Lokrix detects at your URL.

What it is

Optimizing for AI visibility only means something relative to a question. “Am I surfaced?” is incomplete; “Am I surfaced for best GEO platform for enterprise?” is measurable. The prompt universe is that structured question set — a tree of intents (comparison, recommendation, how-to, alternatives, and so on) with the concrete prompts under each.

The prompt universe, branched by intent and stageA root prompt-universe node branches into three categories — intent, funnel stage and type — and each category branches into its leaf values such as informational, commercial and transactional, or branded and unbranded.Prompt universeIntentFunnel stageTypeinformationalcommercialtransactionalTOFUMOFUBOFUbrandedunbranded

How it is generated

When you enter a URL during onboarding, Lokrix creates the property, detects the brand and industry, and generates a prompt universe of the questions buyers in that space actually put to AI engines. It groups them by intent so you can see coverage at a glance and spot where you are invisible.

Its role in scoring and optimization

Each prompt is run through the engines K times to produce a Monte-Carlo presence probability, and those roll up into your Lokrix AI Score and per-prompt share-of-voice. The universe is also what optimization targets: a path to #1 is always anchored to one of its prompts.

Prompts and cost

Live cost is multiplicative — prompts × engines × K samples × cadence — so the size of your universe directly drives credit usage. You choose which prompts to run live and which to leave to the predictive model; see scan cost and credits.