Category Visibility Intelligence is NovaLab's tracking service for brands that need to know how often AI assistants recommend them. Several hundred category prompts run on a schedule across six assistants, multiple times each. Onboarding and calibration costs $6,500. Ongoing tracking costs $3,000 a month.
What you get
Most AI visibility reporting asks an assistant a handful of questions once and calls the answer data. Assistants are non-deterministic, so that number is noise. Category Visibility Intelligence measures a distribution instead, across enough prompts and enough runs for the movement to mean something.
- Category prompt panel. Several hundred questions written for your category, covering what buyers ask at each stage, from category definition through to direct comparisons between named competitors.
- Six assistants, queried separately. Every prompt run across ChatGPT, Claude, Gemini, Perplexity, Microsoft Copilot and Google AI Overviews, because each retrieves from a different index and cites different sources.
- Multiple runs per prompt. Each question asked repeatedly on a fixed schedule, so the result is a distribution rather than one non-deterministic answer.
- Four things captured per run. Whether you appear, where you sit in the answer, how you are framed, and which sources were cited instead.
- Entity resolution. Results resolved against entities rather than string matches, so mentions of your brand, your product names and your people all land on the right record.
- Live dashboard. A URL, not a monthly PDF. Week-over-week movement, competitor share gained and lost, and the specific pages that caused each change.
Pricing
| Tier | What it is | Price |
|---|---|---|
| Onboarding and calibration | Prompt panel built for your category, entity resolution, run schedule set, baseline calibrated. | $6,500 |
| Ongoing | Scheduled runs across all six assistants, multiple runs per prompt, live dashboard. | $3,000 / month |
Onboarding and calibration is a one-off and comes first. The panel has to exist and the baseline has to settle before a monthly number means anything. There is no credit arrangement on this service, because the onboarding work is not an audit that gets superseded. It is the instrument itself.
How it works
- Panel construction, two to three weeks. NovaLab writes several hundred category questions, agrees the competitor set with you, and resolves the entities that results will attach to.
- Calibration. The panel runs against the baseline repeatedly until the variance is understood. This is what separates a real movement from assistant noise later on.
- Scheduled runs. Every prompt runs across all six assistants on a fixed schedule, multiple times per prompt, with each run recorded in full rather than summarised.
- Dashboard. Results land in a live dashboard. Share of answers, position, framing, and the cited sources that appeared in your place, all trended over time.
- Review. A monthly working session on what moved, which competitor gained, and which pages are being cited instead of yours.
Those durations are scoping estimates, not commitments. A category with a large competitor set or an unusual buying process takes longer to build a panel for.
Is this right for you
It suits you if
- You already know AI assistants recommend somebody in your category and want to know how often it is you.
- You have a competitor set you can name and a category with real search demand.
- You are spending on content and need to see whether any of it changed what assistants say.
- You want the measurement to survive an assistant changing its retrieval behaviour.
It does not suit you if
- You want a single score. This produces distributions, and a single score would be a lie about them.
- Your category is too narrow for several hundred distinct buyer questions.
- You need the fixes made, not measured. Start with AI and Search Visibility instead.
- You want to try it for a month. Calibration alone takes longer than that to be worth reading.
Questions people ask
How much does AI visibility tracking cost?
Onboarding and calibration costs $6,500 as a one-off. Ongoing tracking is $3,000 a month. The onboarding fee covers building the prompt panel for your category, resolving the entities, setting the run schedule and calibrating the baseline. The monthly fee covers every scheduled run across all six assistants and access to the live dashboard.
Which AI assistants does the panel run on?
The panel runs across ChatGPT, Claude, Gemini, Perplexity, Microsoft Copilot and Google AI Overviews. Each assistant is queried separately, because they retrieve from different indexes and cite different sources. A brand that appears in Perplexity may be absent from ChatGPT on the same question, and Category Visibility Intelligence reports that split rather than averaging it away.
Why run the same prompt more than once?
Because assistants are non-deterministic. Ask the same question twice and you can get two different answers with two different sets of cited sources. A single run tells you what happened once, which is close to useless. Multiple runs per prompt turn the result into a distribution, so you can see how often you appear rather than whether you appeared on one occasion.
How many prompts are in the panel?
Several hundred, written for your category rather than picked from a generic list. They cover the questions buyers actually ask at each stage: what a category is, how to choose, what things cost, which providers to consider, and direct comparisons between named competitors. The panel is built during the $6,500 onboarding and calibration step.
What does each run actually capture?
Four things. Whether your brand appears in the answer. Where it sits in the answer. How it is framed, meaning the words the assistant uses about you. And which sources were cited instead, by domain and by specific page. That last one is the actionable part, because it names the pages taking the position you want.
Do I get a dashboard or a monthly report?
A live dashboard URL. Category Visibility Intelligence does not ship a monthly PDF. Results resolve against entities and persist over time, so you can open the dashboard on any day and see week-over-week movement, which competitors gained or lost share, and which specific pages caused the change.
What is the difference between this and the Visibility Snapshot?
Scale and continuity. The $400 Visibility Snapshot runs a small prompt panel once and tells you where you stand today. Category Visibility Intelligence runs several hundred prompts on a fixed schedule, with multiple runs each, and keeps the history. The Snapshot is a photograph. This is a time series.
Can you make my brand appear in AI assistant answers?
No, and anybody promising that is selling something they do not control. Assistants choose their own sources and change retrieval behaviour without notice. What Category Visibility Intelligence does is measure the gap precisely and name the pages currently being cited in your place, so remediation work is aimed at evidence rather than at a hunch.
Related services
Category Visibility Intelligence measures. Two other NovaLab services act on what it finds. AI and Search Visibility does the entity, schema and citation surface work that the dashboard points at, and its $400 Visibility Snapshot is the cheaper way to see whether this scale of tracking is justified yet. Measurement and Tracking makes sure your own analytics can corroborate a change. All five sit on the services overview, with every price on the pricing page.
What an assistant answer measures, and why a single check tells you nothing, is set out in the what is AI visibility guide.
Start with a Visibility Snapshot
$400. Before committing to $6,500 of panel construction, run the small version. A prompt panel across the major assistants showing where you appear today and who gets cited instead. The fee is credited against what follows.
Book the Visibility SnapshotLast updated 14 September 2026 · NovaLab Digital · Dubai, UAE