Guide

Vizible tracks how AI models, AI-powered search, and Google talk about your brand — and what to do when the answer isn't what you'd want. This page is the practical walkthrough. For the exact scoring math, see the methodology page.

Setting up a subject

A "subject" is anything you want tracked — your brand, a competitor, a product, even yourself. Each one runs independently with its own prompts and schedule.

Create a subject

Give it a name and pick a template (Brand, Person, Competitor, Product, Topic, Stock, Policy, or Custom). The template just pre-fills sensible starter prompts — you can edit everything after.

Write your prompts

Prompts are the actual questions sent to AI models on every run — e.g. "What's the best project management tool for small teams?" not "Tell me about [Brand]." Write prompts the way a real prospect would actually ask, not the way you'd search for yourself — that's what determines whether you show up.

Pick your models

Choose which AI models to track in the Models tab (GPT, Claude, Gemini, Grok, etc). Track the models your actual customers are likely using, not every model available — more models costs more credits per run without necessarily telling you anything new.

Set a schedule and run

Runs can be manual ("Run now") or scheduled (daily/weekly). Hit run once to get your first score, then let the schedule keep it current — a single run is a snapshot, the trend line is what actually matters.

Reading your results

Vizible Score

The single composite number (0-100): AEO×50% + GEO×30% + SEO×20%. This is your headline number — everything else explains it.

AEO

Answer Engine Optimisation. Do AI models mention you when asked directly via API? This is the models' raw trained knowledge.

GEO

Generative Engine Optimisation. What real users actually see in Google AI Overviews and Perplexity — a live surface check, not just a model's memory.

SEO

Classic organic search ranking. Still matters — AI answers are trained and grounded on the same web your SEO already targets.

SoV (Share of Voice)

Found in the Compete tab — how often you're mentioned relative to the competitors you've added, across the same prompts.

Score label

Dominant / Strong / Visible / Emerging / Weak / Invisible — a plain-English read of the Vizible Score band, so you don't have to memorise thresholds.

When your score changes

Open the History tab and click any completed run — it expands into exactly which prompts moved the score and why (lost a citation, dropped in rank, stopped being mentioned, etc), not just the raw delta number. If a prompt shows zero mentions, the Prompts tab links straight to Compete so you can see who's showing up instead.

Acting on it

Check Insights

The Insights tab generates specific, prioritized recommendations from your latest run — not generic SEO advice, but tied to your actual gaps (a missing citation, a weak AEO mention, an absent SoV signal against a named competitor).

Check Compete

If you're not mentioned, Compete shows who is. That's usually more actionable than "improve your content" — it tells you who you're actually losing the answer to.

Don't chase single runs

One run can move on response wording alone. Look for a pattern across 3+ runs before treating a drop as a real signal worth acting on.

Full scoring methodology →Back to Dashboard →