FAQ
What gets measured, what the contract delivers and how your data is handled, answered in the order a buying team actually asks. Every price is a public USD list price, and every deliverable arrives fully async.
One Panorama asks 50 questions of 21 engines, 7 times each — 7,350 cells in all, read across the 20 places an answer can show your brand.
A CiteAngle audit comes in three widths. Compass ($1,199) asks 20 questions 7 times each across 21 AI and search engines, and nothing gets skipped. One observation is one question put to one engine, one time. Panorama ($2,599) widens to 50 questions, spread over 20 places an answer can appear: 21 engines give 20 because a search that also returns an AI answer is read as two separate places. One Google search hands back both a results page and an AI Overview.
The problem
Buyers have a long list of questions and the answers sit in different places.
What it costs and what actually arrives are the two answers anyone needs before signing.
How we fix it
Every question here is open.
Nothing hides behind a click. The page runs in three blocks: measurement, then product and contract, then how we handle your data.
What you get
By the end you know what you are buying.
Price, deliverables, delivery time and how to stop are all on this page. Anything left over goes to email and comes back in writing.
Why is this a budget line at all? Because the step where buyers form their shortlist has stopped producing evidence. A buyer describes the problem to an answer engine, reads a reply assembled from pages you did not write, and arrives with the decision largely made. Or never arrives. Analytics starts at that arrival, and a rank tracker reports on the links the answer sat above. Neither can tell you what the reply said, whether you were in it, or who was named instead.
Every CiteAngle reading is sealed with its conditions and read side by side with your competitors, so the answer to "did it change?" is a comparison rather than an opinion. Without a fixed protocol, two readings taken months apart cannot be compared, and a year of work loses its evidence along with its budget.
Then every question a buying team asks, in order.
When the answers are done, the next question is price. See program prices →
20
buyer questions in Compass. Each one is put to 21 AI and search engines, 7 times over, measured by machine and reviewed before delivery
Product specification · $1,199 one-time
50
buyer questions in Panorama. The same 21 engines and 7 runs, plus the async commentary layer
Product specification · $2,599 one-time
30-day
credit window. Your full audit list price comes off the first program payment within thirty days
Radar · Compass · Panorama — all four execution programs
Timing — Today's AI answers can only be measured today. Every audit goes into our measurement ledger exactly as measured, the moment it runs — a baseline no one can recreate after the fact — and we keep extending that record, run after run, on the same protocol.
Protocol — We ask the same question seven times, report the confidence interval, and attach a reproducibility receipt carrying the timestamp and configuration hash.
Judgement — We publish the state of every observation — observed, cited, or indeterminate — so every number comes with its evidence. Each citation row carries the source domain and, where the answer exposed one, the full page address, next to the question that produced it and the run it came from.
AI search visibility is a metric that quantifies, through repeated measurement, how often a brand is cited or mentioned in the answers of chatbots and AI search engines. In paid audits, CiteAngle measures it with seven repeated runs (k7) per query and reports every applicable rate with a Wilson 95% confidence interval.
Knowing exactly where you stand is the first thing CiteAngle gives you. Two facts land on the opening screen: which AI surfaces come back with no AI answer shown, our term for an answer that does not carry your name yet, and which brand occupies that spot right now.
The second one carries most of the value. Competitors are counted inside the same answer to the same question, so you also see which document got pulled in to fill the space. That is where the fix list comes from. The free check hands you that first column at no cost, and a paid CiteAngle audit puts the same question up seven times to measure how much the answer moves.
Pick your field and the questions fill themselves in, the ones buyers in that field actually type. The list holds 25 entries today, 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, and so on.
Not seeing your field is fine. Type it into the industry box and CiteAngle builds the questions out of your own words. A funeral home, a photo studio, the method behind them does not change. Which one fits narrows the starting point on one screen.
Google describes optimizing for generative AI search as an extension of search-experience optimization, that is, of SEO. CiteAngle builds on technical SEO, people-first original content and entity fact consistency, then directly observes what AEO and GEO are named for: answer-level brand mentions and citations, plus AI Overviews (AIO) exposure, one integrated visibility read across search and AI answers. Google's official guide
Different labels, same surface. LLMO names the work of getting your brand into what a language model writes. AEO points at the answer itself. GEO points at whether your pages get pulled in as source material while that answer is assembled. All three end up watching one screen, the answer an AI hands back. CiteAngle sorts by screen rather than by label: which engine, which answer, whether your name appeared, which URL was credited as the source. The competitors you name are read on that same screen, in the same run.
Yes. The AI summary that sits at the top of US Google results (AIO · AI Overviews) is judged as its own place: whether it appeared for the query, whether the answer names your brand, and whether your page is cited as a source. It is included from Compass up. Where a client property has Search Console's generative-AI report, we link that export as client-owned first-party evidence, and CiteAngle keeps AIO as an external observed place, measured with repeated runs.
Use both — they answer different questions. Search Console opened its generative-AI report in June 2026: impression-centric exposure inside Google surfaces for a site you own, rolling out progressively. That report is your first-party record, and where a client property has it, we link the export as evidence beside the paid read. CiteAngle measures the answer layer itself, across engines a single console does not cover: whether ChatGPT, Perplexity, Claude, Gemini, Google AI Overviews and Bing Copilot name your brand and cite your pages, with competitors read in the same answers, repeated runs, and Wilson confidence intervals on every applicable rate.
Yes. On queries where an AI summary appears, part of the attention your ranking used to earn moves into the answer itself, and the rank table does not tell you whether that answer carries your brand. SEO remains the foundation. What the report adds is the crosswalk: search presence and AI presence measured on the same queries, in the same round, because some search work demonstrably transfers into AI answers and some does not, and the crosswalk shows which move comes first for your category.
Because the engines disagree. In our measured runs, a brand cited on one engine routinely fails to appear on another. Our US rail follows how US buyers actually search: as of June 2026, StatCounter puts Google at 86.67% of US search and Bing at 8.73%, and Bing carries an AI answer of its own, with Yahoo Search at 2.55% and DuckDuckGo at 1.53% kept as secondary diagnostics. And we don't stop at AI answers: every audit tier runs the same questions across organic search, news, YouTube, short-form and Bing Copilot, 20 places an answer can appear, counted on 21 engine captures, so you can see whether your search equity is transferring into AI answers.
Every run is sealed: we fingerprint the conditions and the raw data when it ends, so any later edit shows. Provenance travels with it, from interface, model and location to capture time, run index and a source receipt, so the same conditions can be measured again. Judgment is separated from collection: citation is verified against the source (deterministic), while mention and sentiment are judged on their own track with labeled confidence levels. Entity disambiguation keeps a same-name company from being counted as your brand. The measurement standards are published in the methodology.
We ask the same question seven times a round and hand you the spread as a number. AI answers do move between runs, and that movement is one of the things the audit measures. If you asked a model twice and got two different answers, you read that correctly: one run is one run, and a single reading is a value, not a trend. Every paid audit reads citation and mention counts inside the collected runs, alongside the answer variation itself. Applicable citation rates carry Wilson score 95% confidence intervals, reported with the effective sample size, and precision labels are granted by the actual width of the pooled interval. The final report is released after the repeated runs and the delivery-package review complete.
Because the ad slot and the answer are different products. OpenAI, Google, and other platforms state in their own documentation that ads are labeled placements around the answer, and advertisers "have no ability to shape, rank, or alter" the answer itself (OpenAI Help Center, verified 2026-07-18). Ad exposure stops the moment you stop paying; presence inside the answer body persists after spend stops (it does require upkeep). And even if you do run ads, incrementality is the question: buying ads on queries where you're already cited means paying twice for reach you already have. That's why a citation baseline, measured across the places AI answers appear, by an independent third party that doesn't sell ads, with point-in-time evidence sealed at capture, is the data that should precede the ad decision, not replace it. Read the full analysis.
Much of what a voice assistant says out loud is produced by text answer engines we already measure. Samsung's Galaxy Bixby serves web-question answers as Perplexity-based search summaries with source cards (One UI 8.5), Apple has been reported to have agreed to bring Google's Gemini to Siri (CNBC and TechCrunch, January 2026), and ChatGPT voice conversations are answered by GPT itself. Perplexity and GPT are judged places CiteAngle measures with seven repeated runs per query, so a large share of "which brand gets cited in the spoken answer" reduces to the citation coordinates of those text engines. What stays voice-specific, from engines still mid-rollout like the Siri rebuild to spoken-answer truncation and device-layer localization, we track as monitored items until they settle.
The planned measurement, the report, the agreed execution and matched remeasurement under the same conditions: that is the contractual delivery standard. In the report that means the source domains and page addresses the answers cited, the competitors counted beside you on the same questions, and a machine-readable JSON file carrying every citation address, whether each engine cited you or named you, and the seven runs one by one. We report the observed change in search and AI visibility together with client-owned conversion evidence, Search Console, Bing and GA4 numbers you can recompute in your own accounts, so the next round of priorities is set by data.
Same measurement engines, same judging rules. What changes is measurement width and interpretation depth.
Radar ($399) puts 5 buyer questions to every place on our US list and asks each one seven times. It settles the first question: does AI know your brand at all, and which sources does it lean on when it answers? It is the short entry read, and its list price comes off a program payment within thirty days exactly like the two below.
Compass ($1,199) widens that to 20 questions on the same 20 places, the grid run in full with every status shown (six AI answer engines + Google AI Overviews + organic search cross-check). Measurement runs automatically, we review the package, and it lands on your dashboard in one to three business days after measurement starts: the SEO×AI quadrant crosswalk, a technical gate audit, the AI brand-recognition read and 3 verified competitor profiles.
Panorama ($2,599) takes it to 50 questions on the full list, adding News, YouTube, Google short-form, Bing and Copilot. It adds 10 competitors, SNS presence cards, the misinformation scan, Wilson score 95% confidence intervals, the completeness manifest that checks the deliverable set for gaps, a narrative read, the 15-slide executive brief and email Q&A with a two-business-day written response. If the result is going in front of a board, it's Panorama.
Every list price, add-on and payment term sits on the pricing page, which is the source these figures are copied from.
Both audits repeat every query seven times per round and release the final report after the delivery-package review, one to three business days after measurement starts. Whichever audit you start with, its full list price comes off the first payment when you begin any program within thirty days: your audit spend becomes execution budget.
They are different classes of measurement.
The free AI visibility check, requested with your scope through the contact page and started the same business day, is a preliminary automated read of roughly 18 observations, taken on one to three buyer questions across six places where AI answers show up. It comes back with the source addresses those answers cited, so you can see where AI places your brand today and what the delivered format looks like on your own name.
Paid audits answer the questions that follow, with numbers: Compass runs 20 buyer questions and Panorama 50, adding full status disclosure and statistics, the technical gate audit, the AI brand-recognition read, win/loss and kickoff cards, citation gaps, the SEO×AI quadrant and entry priorities. Panorama layers on the misinformation scan, the narrative read and the 15-slide executive brief. Compare the real deliverables on the sample reports.
Yes. That is the Refit Sprint ($15,900 · four weeks · SEO·AEO·GEO·AIO). It runs on top of a paid audit: Compass or Panorama fixes the baseline and the priority targets first, then the sprint executes six priority moves: four core pages implemented or patched, two new evidence assets, up to ten technical gates across crawl, indexing, rendering, sitemap and canonical, and the entity fact sheet.
Search Console, Bing AI Performance and GA4 conversion evidence are connected along the way, and the four weeks close with a change-receipt ledger of before/after URLs, a final k7 matched remeasurement and an executive delta memo.
The Refit Sprint ($15,900 · four weeks) implements the priority winning moves your audit confirmed: core pages implemented or patched, new evidence assets, technical gate patches and an entity fact sheet, closing with a change-receipt ledger, matched remeasurement and an executive delta memo. The Category Sprint ($29,900 · eight weeks) widens to two implementation waves with baseline, mid and final measurement, registered held-out and delayed evaluation, an authority-source pack and weekly written decision memos.
Change the core pages and the technical foundation first: four weeks. Build the structure that gets you selected across the category, including the authority-source layer: eight weeks. Both require a paid audit first; that is where the priorities come from.
The contract fixes what lands each month. The commentary layer completes in writing on every program: written memos, recorded briefings and email Q&A with a two-business-day written response. Each program re-measures on Compass or Panorama every month, and every decision is reached without a meeting.
Pay by the month, or once for the year. The yearly price is one payment covering all twelve months, and it costs 2 months less than twelve monthly payments; same work, same monthly rhythm, the invoice is the only difference.
Competitive Signal ($3,490 a month, or $34,900 a year) — every month runs a wide-angle Panorama read across the full grid, with thirty competitors judged in the same answers and month-over-month movement flagged. You also get role, source and misinformation change cards, each with source URLs and run indices, and a written signal memo. The month closes with a delta report, a reading through that month's lens, the sealed evidence ledger and your claims registry entry, yours to keep. Observation-led: your team executes.
Growth Operations ($8,900 a month, or $89,000 a year) — every month a wide-angle Panorama (the full grid, thirty competitors and a month-over-month delta view), a written decision memo, two priority winning moves executed with work orders and completion receipts, and a recorded briefing. It closes with a delta report on matched conditions, the receipt ledger, client-owned conversion evidence read side by side, and next month's priorities. The flagship program.
Category Leadership ($15,900 a month, or $159,000 a year) — adds four executed moves a month, a monthly authority-source campaign pack, a thirty-competitor position panel and a board-ready decision deck every month.
All three are contracted engagements. A person works on your account every month, so each one starts from a written agreement: get in touch and we scope it with you first.
The commentary layer completes in writing. On Panorama you get a narrative read and the 15-slide executive brief; both come out of your own run, and a person signs off before it reaches you. Programs add a monthly written decision memo (PDF) and a recorded briefing; email Q&A answers follow-ups with a two-business-day written response. The evidence, the recommendation and the next order of work arrive as documents and video: you read and watch on your own schedule, in any time zone.
Executive reporting, cross-team alignment and board sessions run the same way. On Panorama and on every program, no meeting is required to reach the decision.
Within thirty days of a paid audit payment, Radar, Compass or Panorama, the full list price of your most recent audit comes off your first program payment. All four execution programs carry it: the Refit Sprint, the Category Sprint, Growth Operations and Category Leadership. Your audit spend becomes execution budget in full.
Treat the delivered coordinates as the starting point. The report prioritizes the questions no one has claimed yet across search and AI, the competitive edges worth pressing and the state of your conversion linkage. Matched remeasurement under the same conditions then shows the change as a delta, and client-owned conversion evidence, Search Console, Bing and GA4 numbers you can recompute in your own accounts, sets the investment order for the next quarter.
Paid measurement raw data, reports and change receipts are handed over within the contract scope (free-snapshot raw data ages out after ninety days). The monthly close reads that month's observation, execution and conversion evidence side by side, and next month's priorities and the renewal scope are decided on that record, which you keep either way.
CiteAngle does not bill you automatically today. Every following term is confirmed by written approval on a fresh quote, and a free service never converts into a paid subscription on its own.
Once online subscription checkout is active, a subscription you order there renews at the listed price for the term you selected, and the renewal date, the renewal amount, how to cancel and when cancellation takes effect are all stated at checkout and in the order record. To stop, email support@citeangle.com before the renewal date or use the subscription-management link that comes with your order; it takes effect at the end of the term you already paid for. The clauses themselves sit in Cancellation and refunds §8 and Terms §5.
We collect the minimum, email, brand and queries, and process it under the US Privacy Notice. Benchmark statistics are built only as anonymous, industry-level aggregates; brand names, domains and other identifying details are not included.
Audits, add-ons and subscriptions carry fixed public US-dollar list prices. The price on the services page is the price you pay, with the payment rail and the refund route stated before any money moves. Invoicing documents are settled before payment, measurement starts once payment is confirmed, and a scope is fully refundable any time before measurement begins. Programs and Enterprise engagements are scoped by written proposal, which fixes the USD price and terms before you pay. Full terms live in the Terms and the Refund Policy.
Paid audits publish to your dashboard one to three business days after measurement starts, and Panorama ships with the 15-slide executive brief.
Each question on its own page
The questions buyers ask last, answered before you pay.
The core is the measurement itself — every planned observation reported with its status — plus citation and mention tables for each place, question by question, each citation row carrying the source domain and, where the answer exposed one, the full page address.
What sits on top of the measurement
On top of that sit the SEO×AI quadrant crosswalk, the competitor read and the prioritized next actions; Panorama adds win/loss and entry cards, the misinformation scan, the narrative read and the executive brief. Reports land on your dashboard and download as PDF, and a machine-readable JSON file travels with them, holding every citation address, whether each engine cited you or named you, and the seven runs one by one. The delivered screens are shown in their real structure on the samples page.
Your brand name and aliases, primary domain, industry, key competitors and the questions your buyers would ask an AI — plus the target state or Washington, DC.
How the panel gets locked
The quote fixes your USD scope first; after payment confirmation you receive a query-panel draft to approve, and the approved question set is version-locked so every later remeasurement compares against the same panel. The more owned web and social assets you list, the sharper the measurement gets.
On your dashboard, typically one to three business days after measurement starts — viewable online, downloadable as PDF, shareable with your team and executives through a report share link valid for seven days, with an email notice when the report is ready.
History, share links and the commentary layer
Your account keeps the measurement history, so repeated engagements read as a time series. From Panorama up, the async commentary layer arrives with it: Panorama's narrative read and 15-slide executive brief, monthly decision memos and recorded briefings on programs, and email Q&A with a two-business-day written response.
The report turns a low score into an ordered work list — exact target URLs or new-page briefs, owners and acceptance criteria, carried out by one of the two implementation sprints.
Match the cadence of your decisions, not a fixed calendar.
Why cadence follows decisions
AI answers move day to day — our own repeated runs and published external experiments observe the same thing — so one audit is a coordinate, and change verdicts need repeated waves on the same protocol. After implementation, the sprint closes with a matched remeasurement; for quarterly budget decisions, the programs add a monthly wave on the same questions. And a baseline cannot be created retroactively — the earlier your first run, the longer your time series.
Every list price, and which page carries it, is on the pricing page: the four audit tiers, the two implementation sprints, the audit credit and what you can check before you pay.
How a program month runs — Fresh measurement on the locked panel · Written decision memo on what to do · Priority moves executed with receipts · Delta report read on matched conditions. every month · same panel · same judging rules.
How an Enterprise engagement is set up — Scope: brands, markets, languages · Term, governance and invoicing in writing · Annual measurement on a fixed panel · Executive reporting and approval workflow. scoped and contracted by quote, before it starts.
The three verification gates — V1 · Baseline sealed before work starts · V2 · Every executed change carries a receipt · V3 · Matched remeasurement at the close. sealed ledger · change receipts · matched rerun.
Five checks, in buying order — Free check on your own brand · Sample reports · Published methodology · Public claims registry · Hash-sealed receipts on delivery. every one of them works before any money moves.
From order to report — Scope and USD quote fixed · Query panel drafted, then you approve it · The measurement runs in full · Report on your dashboard, 1–3 business days. the order of events after payment confirmation.
What comes back after you send this form — You send the scope and your contact · We write the quote in USD · Contract and invoice follow · Measurement starts on your approval. a written quote and contract, in US dollars.
The async commentary layer, named
Async evidence-linked analysis — a named deliverable
A narrative read and the 15-slide executive brief on Panorama; a monthly written decision memo and recorded briefing on every program; email Q&A with a two-business-day response SLA throughout — delivered entirely in writing on your schedule. The read and the brief come out of your own run, and a person checks the package before we publish it. The calendar stays yours.
Panorama and every program complete without a meeting.
The five checks, in buying order
Ready to turn US visibility evidence into a market plan? Tell us the US buyer, target state and decision date.
Request a written proposal