NewMeasure your brand across four AI assistants

Do AI assistants recommend you?

Buyers now ask ChatGPT, Claude, Gemini and Perplexity what to buy. Citerva asks the same questions on your behalf and measures how often you are named, cited, and beaten — with the evidence behind every number.

See the plansHow it works

Pro and Growth each start with a 3-day trial, card required.

Example project

Applicant tracking system

Illustrative

Mention rate

30%

11 of 37 runs

Share of voice

13%

11 of 84

Citation rate

17%

3 of 18 runs

Coverage

100%

10 of 10

By provider

OpenAI30% 3/10
Anthropic44% 4/9
Gemini0% 0/8
Perplexity40% 4/10
Top competitor named in 86% of the same answers4 of 48 runs did not answer and are excluded from every rate
Measurable visibility

Four questions, answered with numbers. Not a score out of a hundred. Rates over denominators you can see, from answers you can read.

Are we visible?
Mention rate across every prompt and provider — the share of answers that named you, out of the answers that actually came back.
Who beats us?
Competitor mention rate and share of voice, scored over exactly the same answers as you, so the comparison is like for like.
Where are we absent?
The specific questions where an assistant answered, named your competitors, and did not name you.
What should we investigate?
Evidence-linked findings that point at what was observed — never a promise that a change will move a model.

Honest by construction

A number you can check

Most of this category reports a score and asks you to trust it. Every figure here comes with the answers it was read from.

A provider we did not ask is a gap, not a zero
If an assistant times out or has no API key configured, that run is excluded from every rate and shown as a gap. Reporting it as "did not mention you" would be a false claim about a third party.
Every raw answer is kept
We store what each model actually said, verbatim. When our parser improves, past scans are recalculated from the original answers instead of history quietly changing meaning.
One scan is a sample, not a ranking
AI answers vary between runs. We show the counts behind every rate, flag small samples, and refuse to call a one-run difference a trend.

How it works

From a product description to a measured answer

Five steps. You approve every question before it is asked.

  1. 1

    Describe your product and your competitors

    Plus the topics a buyer would actually be trying to solve.

  2. 2

    Review the generated prompt set

    Real buyer questions — "best applicant tracking system for a 20-person startup", "alternatives to X". Edit or reject any of them. Nothing is sent until you approve it.

  3. 3

    Run the scan

    Every approved prompt is asked of every provider you enabled, and the full answer is stored.

  4. 4

    Read the evidence

    Mentions, competitors and cited sources, parsed from each answer and traceable back to the text they came from.

  5. 5

    Repeat on a schedule

    Weekly or monthly on Pro, daily on Growth, so you are reading a trend rather than a single sample.

Pricing

Priced on what actually costs money

Every scan spends real money with AI providers, so plans are sized by provider calls rather than by feature checkboxes.

Pro
Scheduled tracking across all four AI providers.
$49.00
Billed monthly
  • 3-day trial, card required
  • 3 projects
  • 50 prompts per scan
  • All 4 AI providers
  • Up to 8 scans per month
  • 1,000 provider calls per month
  • Weekly and monthly scheduled scans
  • 5 team members
  • Mention rate, share of voice and citation analysis
  • Prompt-level evidence and recommendations
Growth
More projects, larger prompt sets and daily schedules.
$149.00
Billed monthly
  • 3-day trial, card required
  • Everything in Pro
  • 10 projects
  • 150 prompts per scan
  • All 4 AI providers
  • Up to 40 scans per month
  • 6,000 provider calls per month
  • Daily, weekly and monthly scheduled scans
  • 20 team members

Pro and Growth each start with a 3-day trial, card required. Online checkout for Pro and Growth is not open yet.