[ AI search visibility, measured honestly ]

When someone asks an AI who to hire, is it your name that comes back?

SparkBeacon asks ChatGPT, Claude, Perplexity and Google's AI Overviews your customers' real questions, repeatedly, and reports how often each one names you, with the uncertainty, per engine, and what to do about it.

See the four rungs
SparkBeacon · LiveAnne Sostman · through 2026-09-21
3.3%
ChatGPT
6.1%
Claude
8.8%
Perplexity
5.6%
Google AI Overview

Named in AI answers, per engine and never averaged. 983 answers, 4 runs pooled over 28 days; the 95% intervals are in the table below.

Questions16 on the panel · 6 invisible · 3 cited · 7 named · 0 recommended
Per engineChatGPT 3.3% (1.7% to 6.3%) · Claude 6.1% (3.7% to 9.8%) · Perplexity 8.8% (6.1% to 13%) · Google AI Overview 5.6% (3.1% to 9.7%)
Sources96% earned · video #7 on Google's answers
Movement0 confirmed · 3 too early to call
Not a sample. The founder's own dashboard, same ledger, refreshed hourly.
[ The model ]

Four rungs. Different causes. Different fixes.

Being cited, being mentioned, and being recommended are three different events with three different owners. Most tools measure the second and sell it as the fourth. SparkBeacon measures each rung separately, because the work to climb each one is different.

1

Retrieved

Entirely ours

A crawler reaches and parses the page. If the firewall turns it away, nothing below this line can happen.

SparkBeacon knocks on the door as each AI crawler, and a reporter on the brand's own server logs the visits that identify themselves as those crawlers. Read the log as a floor: names are not yet checked against the engines' address ranges.
2

Cited

Mostly ours

An engine uses the page as a source and links it. The page was good enough to quote.

Every source is kept per question and per engine, marked as the brand's own or somebody else's.
3

Mentioned

Our copy decides

The brand name appears inside the sentence the reader sees. The cheapest rung to fix and the most often skipped.

Every answer is read for the name, aliases and misspellings included. A question quoted back does not count.
4

Recommended

The rest of the web decides

The model shortlists the brand for a stranger's unbranded question. Earned sources put a name there.

Recommended means named in at least half the answers and described favourably. One lucky answer is not a recommendation.

Recognition is not recommendation. In one test of 112 startups, 99.4% were recognised when named; 3.32% surfaced in unbranded discovery.Reverse-engineered

Single-author thesis (arXiv 2601.00912), not peer-reviewed. Direction matches our own ledgers; the exact figures are one sample.

Cited is not mentioned. One of our own brands: cited six times, named zero, competitor recommended in the same answer. The fix was a sentence.Our data

SparkBeacon ledger

The mention is where the reader stops. Click-through 8% with an AI summary, 15% without; 1% clicked a link inside the summary.Proven

Pew · 68,879 searches · 900 US adults · March 2025

[ How the instrument is built ]

One layer per rung, so a low number always says why.

A single visibility score hides four different problems. SparkBeacon is structured so that every question sits on a rung and every rung has its own evidence and its own next step.

Rung 1 · Retrieved

Can the crawlers get in at all?

The site audit presents each AI crawler's own user-agent and records whether the firewall serves it. A reporter on the brand's server then logs the visits that identify themselves as those crawlers, by name and not yet by address, so the log is a floor. Where a firewall verifies bots by network address, the audit says unverifiable instead of guessing, because a request wearing a crawler's name from anywhere else is supposed to be refused.

Site audit · crawler accesslive
GPTBotserved
OAI-SearchBotserved
ChatGPT-Userserved
ClaudeBotserved
Claude-Userserved
Claude-SearchBotserved
PerplexityBotserved
Googlebotserved
Bingbotserved
Applebotserved
Meta AIserved
Bytespiderserved
CCBotserved
Rung 2 · Cited

Which pages did the answers come from?

Every source an engine used is kept and marked as the brand's own or someone else's. A channel or profile on a platform the brand does not own counts as its own page too, so a video library is credited from the first video. The earned share is measured for this brand, not quoted from an industry average.

Sources the AI trustslast 28 days
annesostman.com487×
zillow.com473×
homelight.com463×
fastexpert.com398×
realtor.com341×
homes.com307×
96% earned · video is source #7 on Google's answers (3.2%)
Rung 3 · Mentioned

Is the name inside the sentence?

Each question is asked several times per engine and every answer is read for the brand's name. The result is a rate with a 95% interval, because five answers can differ from the next five by sixty points. The gap between cited and mentioned is copy work: the name goes inside the sentence, never beside it.

Named in answers · per enginenever blended
ChatGPT3.3% · 1.7% to 6.3% · 8 of 245
Claude6.1% · 3.7% to 9.8% · 15 of 246
Perplexity8.8% · 6.1% to 13% · 26 of 294
Google AI Overview5.6% · 3.1% to 9.7% · 11 of 198
Cited, not named111 of 487 times the brand's pages were used
Rung 4 · Recommended

Does a stranger's question bring the name back?

The panel is built from unbranded questions in the customer's own words: who should I list my house with, which agent is best at marketing a luxury home. Who gets named instead is listed, question by question. A rung crossing only counts as a move when the two windows' intervals stop overlapping, in either direction.

Visibility ladder16 live questions
Recommended0
Named7
Cited, not named3
Invisible6
0 confirmed moves · 3 too early to call
[ How we measure ]

The number is only as good as the protocol behind it.

AI answers change from one ask to the next. A tool that asks once and reports a percentage is reporting noise with a decimal point. These are the rules every SparkBeacon figure passes before it is shown.

±72
points, one answer

The 95% interval on a single answer. It cannot tell 10% from 80%.

±30
points, seven answers

Still wide. Recent runs of the same question pool together, and at about 35 answers the interval reaches ±16, without paying for it twice.

91%
of citations, one engine only

A blended score describes an engine that does not exist. Every figure names its engine.

Before a number is shown

  • Seven or more answers per question. Below that, the interval is wider than the claim, so the page says too early.
  • Per engine, never blended. The same brand scores two to three times differently across engines.
  • Brand mentions, not URL citations. About 65% of cited sources change day to day; the name in the sentence is the stable signal.

And after

  • A two-to-four week rolling window. Movement is measured between two windows that do not overlap.
  • Read the raw answers. Whose name is in the sentence, and did the engine search at all? Every stored answer is readable behind the sign-in.
  • Refuse to flatter. A crossing the intervals cannot separate is shown as too early to call, including the ones that would look like our work succeeding.

Intervals: Wilson score, 95%. Study figures from the sources cited in The Visibility Standard v2.1; nearly every large study in this field is vendor-run and the Standard applies its own test to each.

[ The next question ]

The follow-up is the new keyword.

After the first answer, a searcher asks one of five things next. Each is a different page. SparkBeacon maps the customer journey from real search demand and Google's own what-people-ask-next chains, then sorts every follow-up into its type so the missing page is visible before anyone has asked for it out loud.

  • Clarification · 3
  • Constraints · 4
  • Comparison · 2
  • Validation · 4
  • Action · 2

The five types

  • Clarification. An explainer that defines the thing in one paragraph, then the mechanics.
  • Constraints. A costs-and-limits page: real numbers, timelines, eligibility, exceptions.
  • Comparison. A side-by-side that names the trade-offs and says who each option suits.
  • Validation. A proof page: outcomes, reviews, what can go wrong, who it is not for.
  • Action. A do-it page: numbered steps, what to prepare, where to start today.

What a zero means

A type with no live follow-up is a page nobody has written for this journey yet. It is the cheapest page to be first on, because the demand is already visible in the chains and no incumbent is answering it.

Live from Anne Sostman's journey map · sorted by wording, free, deterministic

[ What we do not sell you ]

Things the evidence says not to chase.

Part of understanding a field that changes monthly is knowing which of last quarter's tactics have already been tested and failed. SparkBeacon will not charge you for any of these.

X-1 Refuted

llms.txt as a citation lever

97% of 137,000 domains' files got zero requests (Ahrefs, May 2026); Google does not read it. We generate one because it is free, and say so on the file.

X-2 Refuted

Schema sold for citations

Controlled test: AI Overview citations −4.6%, AI Mode +2.4%, ChatGPT +2.2%. Required as a commerce mechanism, not a citation lever.

X-3 Proven

Blended scores

One number across engines describes no engine. Every SparkBeacon figure names its engine.

X-4 Proven

Self-promotional listicles

69% of the citations they earn appear in answers recommending a competitor.

X-5 Proven

A single run as a percentage

Below seven answers the interval is wider than the claim. We pool, or we say too early.

X-6 Proven

Porting one brand's channel mix

Reddit is 20.8% of the top-50 external citation domains in one B2B software sample and 1.1% of citations in a realtor sample. Different denominators, and a nineteenfold gap. Measured per brand, never inherited.

X-7 Refuted

Bing optimisation for ChatGPT

The API path was retired in August 2025.

X-8 Refuted

AI detectors

OpenAI's caught 26% and false-flagged 9%; 61% false-positive on non-native English. Filler is penalised, not assistance.

[ How a brand is onboarded ]

Eight dials. Everything else is fixed.

The Visibility Standard treats the mechanics as constant and the brand as eight variables. Setting them in order is the whole onboarding; a blank dial is usually the finding.

Dial 1

Entity string

One name, address, phone, licence and domain, byte-identical everywhere.

Dial 2

Category noun

In the customer's words, never the trade's.

Dial 3

Market scope

Where the buyer actually is.

Dial 4

Ideal client

One person, chosen A or B on every dimension. Both is not an answer.

Dial 5

Surface mix

Measured, not assumed: which sources the engines draw on for this brand.

Dial 6

Proof currency

The evidence this market believes: photos, numbers, video, reviews.

Dial 7

Story node

The one true before, from the brand's own past.

Dial 8

Regulatory overlay

Which regulator, which rules, and the named person who decides.

In order: fill the dials · verify rung one · publish the ungated evidence layer · commission the first proof · set the entity string everywhere · baseline under M · only then choose channels · name the compliance owner

[ Everything in the instrument ]

One dashboard, every layer, one ledger.

Every measurement is stored forever, so any before-and-after is a query rather than a memory. These are the modules, each with its own append-only history.

01

AI visibility runs

Your questions, four engines, repeated, with intervals. A stopped run resumes without paying twice.

02

Visibility ladder

Every question on its rung, pooled over a window, with confirmed and too-early moves.

03

Progress report

Baseline against now: what is working, what is not, competitor movement with how and why.

04

Customer journey

Stages from search demand, the follow-up chains, the five follow-up types.

05

Recommended instead of you

Who the engines name on the questions where you never appear.

06

Sources the AI trusts

Cited domains, earned versus owned, where video ranks on your own questions.

07

Crawler access and activity

Firewall answers per crawler, and a server-side reporter for the visits that identify as each crawler.

08

Site audit

Crawl, AI readiness, issues with copy-paste fixes, a free llms.txt.

09

Rank tracker and organic

Measured positions from live results pages, plus the licensed index's view.

10

Listings, map boxes, NAP

Where you are listed, where details differ, which searches show a map box.

11

Competitors and authority

Auto-discovered rivals, why their authority differs, how to overtake it.

12

Ads and search gaps

What runs on your searches; where page one is weak enough to take.

13

First-party traffic

Humans arriving from AI, with the referrer, plus branded-search demand.

14

Usage and limits

What your package includes and where you are this month, on one page.

[ How it will be sold ]

A package with limits you can see, and nothing that runs on its own.

Each package sets how many questions are on the panel, how many answers are sampled per question per engine, and how many runs a month. The limits are on the settings page, month to date, so there is never a surprise. Pricing is not published yet; the launch list hears first.

Panel
questions measured

How many of your customers' questions are asked, across four engines, every run.

Depth
answers per question

How many times each question is asked per engine. More answers, tighter interval. Recent runs pool, so depth compounds.

Cadence
runs per month

Weekly or monthly, plus the ones you press yourself. Every run is kept forever.

Every package includes the full instrument: ladder, journey, sources, crawler access, site audit, rank tracking, listings, competitors, progress report. The package changes how much is measured, never what you can see.

[ Launch list ]

Find out what the engines say when nobody says your name first.

One email when SparkBeacon opens. Nothing else.

SparkBeacon · a Spark Noir property · built to The Visibility Standard v2.1Page reviewed 2026-09-16 · Sign in