AI Visibility Audit
The AI Visibility Audit from Sigzen AI runs 150 buyer prompts across ChatGPT, Gemini, Perplexity, Google AI Mode and Claude and reports your share of answer with confidence intervals against five competitors, a citation-source map, a hallucination and sentiment register, an AI-referral baseline and a 90-day plan, read out ten working days after kick-off.
$3,500₹1.5L/ ₹1.5L$3,500 · one-time2–3 weeks
Audits are credited 50% against a sprint booked within 30 days.
What you get.
Six things, each with a number you can hold us to.
01
Share-of-answer baseline with confidence intervals
150 buyer prompts, drafted by us and approved by you, run across five engines through official APIs. Share of answer is reported per engine and overall, with confidence intervals, so a move is a move and noise is noise.
02
Competitor benchmark
The same prompts scored for five competitors you choose, so you see who owns which prompts today and how far the gap is on each one.
03
Citation-source map
Which domains each engine cites for your category, how often, and whether you are on them. This is the list the off-site work is built from.
04
Hallucination and sentiment register
Every factual error an engine makes about you, with the source it came from and the fix, plus the sentiment of each answer that names you.
05
AI-referral baseline from your analytics
An AI-referral channel group set up in your analytics on day one, with the sessions and conversions it is sending today, so the before is recorded before any work starts.
06
90-day plan with owners
Three prioritised fixes, then a 90-day plan with each action assigned to a named owner on your side or ours, and the sprint options priced so the next step is visible.
What the report looks like.
Seven sections, read out with you in a 60-minute call. Every number carries its confidence interval and its source.
- Summary and the three numbers: One page: share of answer, citation count and factual-error count, with the headline finding and the three fixes. Example: Share of answer 14% (CI 9–19%) across 150 prompts; cited by 2 of 5 engines on average; 3 factual errors found
- Prompt set and method: The 150 prompts by intent group, the five engines, run dates, the scoring version and how the intervals are computed. Example: Group B, comparison: 38 prompts; each run three times per engine; scoring method v1.2; intervals at 95%
- Share of answer by engine and competitor: Your share on each engine against the five competitors, with the intervals, and the prompts where the gap is widest. Example: Perplexity: you 22% (CI 15–29%), competitor A 41% (CI 33–49%); widest gap on 'alternatives to' prompts
- Citation-source map: The domains each engine cites in your category, ranked by frequency, with your presence on each marked. Example: Top sources: a review site (cited on 31 prompts), a community thread (19), an editorial comparison (12); you appear on none
- Hallucination and sentiment register: Each factual error, the engine, the likely source and the correction; sentiment of every answer that names you. Example: Gemini states a pricing tier withdrawn in 2024, sourced from an unclaimed directory entry; correction: claim and update the entry
- AI-referral baseline: Sessions, pages and conversions arriving from AI engines today, from the channel group set up in your analytics. Example: AI-referral channel group created; 12 sessions and 1 enquiry in the trailing 28 days, mostly to the pricing page
- 90-day plan: The prioritised actions, their owners and the measurement cadence, with the sprint options priced. Example: Weeks 1–2: claim and correct three directory entries (owner: client). Weeks 3–6: rewrite top 20 pages to answer first (owner: Sigzen AI)
90-day plan
The prioritised actions, their owners and the measurement cadence, with the sprint options priced.
Weeks 1–2: claim and correct three directory entries (owner: client). Weeks 3–6: rewrite top 20 pages to answer first (owner: Sigzen AI)
Ten working days.
From kick-off to read-out.
01
Day 1
Kick-off: we present the draft prompt set and the five competitors; you approve or amend both. The AI-referral channel group goes into your analytics the same day.
02
Days 2–5
The 150 prompts run across the five engines on our AI answer monitor, each prompt several times per engine, so the intervals are real.
03
Days 6–8
Analysis: competitor benchmark, citation-source map, hallucination and sentiment register, and the 90-day plan with owners.
04
Day 10
A 60-minute read-out with your marketing lead and one decision-maker, then the report and the dashboard access are yours to keep.
05
Within 30 days
Half the audit fee is credited against an AEO/GEO sprint booked within 30 days of the read-out. If your share of answer is already fine, we say so and you keep the report.
Questions about the audit.
We draft them from the prompt-set template for your industry and your buying stages; you approve them at kick-off. Prompts are run in the engines' default modes through official APIs, several times each, so one lucky answer cannot move the number.
Free, in two minutes. Enter your domain and we'll score it against three competitors across five engines.
No account needed. The report is emailed within 24 hours.