What is an AI Scan?

An AI Scan executes a set of prompts against selected AI models and records the response.

Last updated: 2026-05-05

TL;DR — An AI Scan is one execution batch that asks selected prompts to selected AI models and records every response, brand mention, source citation, and token cost. It's the foundational measurement that powers every dashboard widget in Pranas — Visibility, SOV, Top Brands, Co-occurrence, Sources, Perception, and Action Center recommendations are all derived from scan data.

Why AI Scans are the heart of Pranas

Generative AI engines — ChatGPT, Google AI Overviews, Perplexity, Claude, Gemini, Bing Copilot — increasingly answer the questions that used to drive prospects to your website. If your brand isn't mentioned, ranked highly, or correctly positioned in those answers, you're invisible at the moment of decision. AI Scans let you measure that visibility empirically and repeatedly, so you can track trends, spot regressions, and validate that the changes you make to your content and PR strategy actually move the needle.

Without scans, Pranas would be a static dashboard with nothing to display. Every metric you see is a derived view of one or more scans.

What a scan records

For each (prompt × model) execution, an AI Scan stores:

Everything is queryable from the Results page and aggregated across the dashboard.

Terminology note

You'll occasionally see the word "simulation" in older documentation, in URL paths (/results/simulation/...), and in API/database fields. AI Scan is the current user-facing name; "simulation" is the same thing internally. They're interchangeable — we renamed for clarity, didn't change behavior.

How a scan fits into the bigger picture

Brand → Prompts → AI Scan → Results → Dashboard widgets
                                   → Brand Perception
                                   → Sources & Citations
                                   → Action Center recommendations
                                   → Content Studio briefs

A single brand will run dozens or hundreds of scans over its lifetime in Pranas — typically one weekly scan covering 25–50 prompts across 3–5 models. Time-series across those scans is what gives you trend insight.

Worked example

Acme Feedback runs a weekly scan: 30 prompts × 4 models = 120 executions. After completion:

• AI Score updates on the dashboard (with a sparkline showing change vs. previous week).
• SOV recomputes against registered competitors.
• Top Brands refreshes — Acme moved from #5 to #4.
• Co-occurrence shows Acme is now appearing alongside ProductBoard 38% of the time (up from 31%).
• Sources flags TechCrunch's "Best feedback tools 2026" article as the #1 cited source — Acme isn't in it. That becomes an Action Center recommendation: "Pitch TechCrunch to be added to this roundup."

One scan → measurement, comparison, trend, source intel, and a concrete action. That's the workflow loop AI Scans enable.

Who can run scans

• The brand owner (you).
• Any teammate invited to the workspace with appropriate permissions (multi-seat plans).
• Scan execution is blocked until your email is verified and your credit balance is sufficient.

Edge cases & gotchas

• Scans are not free. Each (prompt × model) execution consumes credits at $1 = 1 credit. See Credits per scan.
• Scans are queued, not synchronous. They survive page refreshes — you can launch one, close the tab, come back later.
• Failed queries auto-retry once with a doubled timeout. You don't need to manually press retry.
• No brands found ≠ failure — it's signal. Sometimes the AI answers without naming any brand at all. See Scan statuses explained.
• Scans are deterministic per attempt, not across attempts. AI models are non-deterministic by nature; running the same scan twice will produce slightly different responses. Use multi-prompt, multi-week scans for stable metrics.

Troubleshooting

Related articles

• Run your first AI Scan
• Scan queue and retries
• Choose AI models for a scan
• Scan statuses explained
• Credits per scan
• Reading scan results

Still need help?

Email contact@pranas.co with the scan ID (visible on the Results page).