Industries
Pick your field and the questions fill themselves in — the ones your buyers actually type
25entries sit on the list, and the last one is a blank you write in yourself
- Law firm
- Dental practice
- Med spa & aesthetics
- Medical clinic
- Chiropractic & physical therapy
- Veterinary & pet care
- Real estate
- Home services (HVAC · plumbing · roofing)
- Construction & remodeling
- SaaS & software
- IT services & MSP
- Marketing agency
- Accounting & tax (CPA)
- Financial advisory & insurance
- E-commerce & DTC brands
- B2B manufacturing
- Logistics & 3PL
- Staffing & recruiting
- Education & training
- Fitness & wellness
- Restaurants & hospitality
- Hair salons & beauty
- Hotels & travel
- Auto services & dealerships
- Other
Not seeing your field? Type it in and CiteAngle builds the questions out of your own words. A funeral home, a photo studio — the method does not change.
How to read this page
The question you need answered decides the first purchase, not the budget. Find the paragraph that sounds like your week, then take the one recommendation under it.
What waiting costs. Without a reading, "we are probably fine there" and "we are invisible there" feel identical from the inside. The difference surfaces the day a competitor turns up in every reply.
All five situations start in the same place. An answer screen said something about your category, a buyer read it, and nobody on your side can repeat what it said. Analytics sees the visit after the shortlist is set; a rank tracker sees the links the answer sat on top of. The screen that did the deciding leaves no row in your stack.
Reading it back is arithmetic, and the arithmetic is published. The widest panel puts 50 buyer questions through 21 engines, 7 times each, landing on 20 surfaces. The narrower audits below run the same protocol on fewer questions, so a smaller purchase buys a smaller sample of the same method.
If none of the five reads like your situation, the free AI visibility check returns your own number on your own queries, which usually settles it.
Five situations, five starting points
Situation one — nobody here knows what AI answers say about us
You are here if
The question has been raised in a meeting and nobody could answer it
Someone asked what ChatGPT says when a buyer describes your category, and nobody could answer. With no earlier reading to point at, the room cannot agree on whether to fund this.
The free AI visibility check returns a same-day read on your own brand, so the meeting has a number instead of an opinion. Radar then confirms presence, absence and per-engine variance across a full-width panel — enough to decide whether a wider audit is worth buying. Compare the audits side by side.
Situation two — we rank fine, but a competitor keeps showing up in the answers
You are here if
Your search positions are healthy and the AI answers still name someone else
Rank trackers say you are winning. The assistants your buyers open cite a competitor on the questions that matter, and nobody can say on which ones or how often.
Fifty queries across the full surface set, with ten competitors read on the same panel and source URLs, answer spans and run indices attached — so the gap is located by question and by surface rather than argued. The narrative read from that same run turns it into an ordered move list. What Panorama delivers.
Situation three — the answers state something about us that is wrong
You are here if
A wrong fact keeps reappearing and you cannot tell where it enters
An outdated price, a discontinued product, a former executive, a location you left. You corrected your own site and the answers did not move, because the source feeding them sits somewhere else.
Panorama’s misinformation scan records where the claim appears and which sources carry it, with receipts. The sprint then runs the correction path — entity facts, source outreach, the pages the engines resolve against — and closes with a matched remeasurement on the sealed panel. How the sprint runs.
Situation four — we have to report this to the board every quarter
You are here if
The number has to survive a room that will ask how it was produced
You need a figure you can defend line by line, and a comparison that means something quarter over quarter.
Monthly full measurement on a sealed panel, a written decision memo each month and a quarter-close delta report with Wilson intervals, read next to your own GA4 and Search Console exports — numbers your team can recompute in your own accounts. The measurement cadence.
Situation five — we are an agency and our client is asking about AI visibility
You are here if
You own the client relationship and need measurement you can stand behind
You will not resell a dashboard whose method you cannot explain, and one account does not justify building a measurement practice from scratch.
Measurement depth under your brand relationship, with the protocol published so your client can check it themselves and the receipts ledger transferring under the engagement terms. How partners sell with us.
What the grid reads, and what the execution layer carries
The grid reads what an outside party can observe: AI answer surfaces, classic search, news, video and the technical layer underneath. In-app search, voice assistants and personalized screens sit in the execution layer — worked on with prescriptions bound to the platforms’ own documentation, and labelled that way in the report instead of carrying a measured number.
The full boundary is drawn on the coverage map, and the protocol behind every measured reading is published on the methodology page.
Still deciding?
Run the free AI visibility check on your own brand first. It takes your queries and returns a same-day preliminary read in the report format you would receive. Every list price is public on the pricing page when you are ready.
Whichever situation fits, the reading behind the recommendation works the same way. Your competitors are read inside the same answer, on the same question, in the same run, so you get a position on a field. The conditions each figure was measured under are sealed at measurement time, so a number in your board deck survives being questioned six months later. Readings we planned but could not take are printed as a count, since a denominator that quietly shrinks flatters every rate above it.
Comparing us against another vendor? The nine questions worth asking put both on the same axes, and the cost calculator adds up whatever combination you pick. The company page names who signs off on the numbers.