Which AI engines do you cover?

ChatGPT, Claude, Perplexity, Gemini, plus Google’s AI Overviews and AI Mode — on every paid plan, with no per-engine add-ons. Otterly doesn't track Claude at all, and Peec keeps it behind custom-priced Enterprise.

How often do you check?

Weekly. A single reading can't tell you much, because the tail of AI answers rotates constantly while a small core of names stays put. So we track the trend and tell you which one you're in — and only alert you when a move is bigger than normal drift.

Is there really a free plan?

Yes. You get a one-time batch of free checks — no card — which is enough to see your first Core vs Carousel verdict. There's also a public checker you can run right now without signing up. The free credits don't renew; paid plans refresh every month.

Do you just measure, or help me fix it?

Both. Every plan includes a GEO audit that scores your pages, a prioritised fix list, and blindspot reports showing the prompts where competitors get named and you don't.

◎CoreCited

AI search visibility · GEO · AEO

Does ChatGPT recommend your brand?

Q: How is this different from my SEO tools?

Traditional SEO tools track Google rankings. We track whether AI engines — ChatGPT, Claude, Perplexity, Gemini — actually mention your brand when buyers ask them questions. Ranking #1 on Google no longer means you appear in the AI answer.

Track your brand across six AI surfaces. Find out whether you're in the part of the answer that stays — or the part that rotates out by Friday.

Start free — no card See what the audit returns

20 free checks · no credit card · results in about a minute

Or check a domain right now

Yyourbrand.com

Week 34 · weekly, N=1

Overall visibility

42%▲ 6

Named in 42 of 100 answers

ChatGPT58%

Claude41%

Perplexity67%

Gemini33%

AI Overviews22%

AI Mode30%

ChatGPTClaudePerplexityGeminiAI OverviewsAI Mode

All six on every paid plan. No engine sold separately, none held back for Enterprise.

[01] THE MEASUREMENT

Being mentioned once is luck. Staying is a position.

Ask an AI the same question eight weeks running and you get eight slightly different answers. A tool that checks once and reports a number is measuring the weather. We run every prompt on a fixed weekly cycle and classify each brand by how often it actually survives — because the difference between appearing 8 times in 10 and 3 times in 10 is the difference between a position and a coin flip.

CAROUSEL

named in <40% of cycles

EMERGING

≥40%

CORE

≥70% — a position

Rival D11%

Rival C27%

You46%

Rival B63%

Rival A84%

Your last 8 cyclescarousel → emerging

Named in 46% of cycles, this brand is emerging — climbing, but not yet a position. A tool that only counted mentions would have called week 7 a breakthrough and week 3 a crisis. Neither was either.

86.5%

of prompts keep a stable core of 1–5 names that persist week after week. Those are positions, and they are winnable.

58–62%

of Google searches now end without a click. The answer is on the page, and either your name is in it or it isn't.

38%

of pages ranking #1 on Google get cited by AI. Ranking first stopped being the same thing as showing up.

Volatility cuts both ways, and it is why we do not alert on every movement. A brand that drops from 46% to 41% has almost certainly not done anything wrong. We tell normal drift from a real change, and only send you the second kind.

[02] WHAT YOU SEE

Every score traces back to the answer it came from.

No black-box index, no proprietary score nobody can audit. We store the actual answer each engine gave, mark where your brand appears in it, and the number you see is derived from those stored answers. If you disagree with a score, you can go read the thing it was computed from.

Share of voice

Across 40 prompts, with 90% intervals.

Rival A31% (25–37)

Rival B24% (19–30)

You18% (13–24)

Rival C15% (11–20)

Rival D12% (8–17)

Blindspots12 prompts

Rivals named. You are not.

“best ats resume checker for tech roles”

named: Rival A, Rival B

“free resume scanner that shows the score”

named: Rival A

“how do I know if my CV passes an ATS”

named: Rival B, Rival C

“resume tools recruiters actually respect”

named: Rival A, Rival D

The answer we stored

Perplexity · cycle 34 · 14 Sep

For applicant tracking compatibility, most people land on Rival A or Rival B. Yourbrand is worth a look if you want the score broken down by section rather than a single number…

position 3rdsentiment positivecited 2 sources

Per-prompt, per-engine

Not one blended number. Perplexity naming you and Gemini ignoring you are two different problems with two different fixes.

Competitors, side by side

Share of voice with an interval attached, because weekly N=1 sampling is not a precision instrument and we will not pretend otherwise.

Blindspots

The prompts where your rivals get named and you do not. This is the list that turns into work.

Sentiment and position

Being named third in a hedged sentence is not the same as being the recommendation.

[03] THE AUDIT

31 checks per page. Every failure becomes a named job.

Knowing you are invisible is not useful on its own. The audit crawls your pages and runs the same 31 deterministic checks on each — no model, no judgement, no score that moves when you re-run it. Everything that fails comes out the other side as a task with a URL, an impact and an effort estimate, so the output is a list of jobs rather than a list of opinions.

Site score

Median across 24 pages, 31 checks each. Deterministic — the same HTML scores the same every time, so a change in the number means a change on the page.

Crawlability

Answerability

Structured data

Performance

Entity

Freshness

Content

A check that cannot apply to your page is removed from the total, not failed. A blog post is not marked down for having no pricing schema.

Tasks from this audit69 open · 12 done

No FAQPage schema — the section is there, the markup is not

/pricingeffort M

high

353 characters before the page answers its own headline

/blog/ats-guideeffort S

high

0 of 6 headings are phrased as questions

/featureseffort S

med

✓

No machine-readable published date

/abouteffort S

med

Title is 71 characters — truncates in the SERP

/integrationseffort S

low

Site-wide findings

Things no single page can show you — found by comparing all 24.

Duplicate title

Four pages share “Pricing — Yourbrand”

Duplicate description

The same meta on every /blog/tag/ page

Thin content

Under 300 words — nothing there to cite

Broken link

Internal links returning 404

Orphan page

Nothing on the site links to it

Redirect chain

Two hops before a 200

These sit outside the 0–100 score on purpose. Folding them in would move the number without anything on the page having changed.

What the 31 checks cover

Crawlability

robots.txt, llms.txt, canonical, indexability, sitemap presence

Answerability

Does the page answer its own headline in the first paragraph, in question-shaped sections

Structured data

FAQPage, Article, Product, Organization — present, valid, and matching the visible page

Performance

Render-blocking weight, image sizing, time to first byte

Entity

Whether the page states who you are in terms a model can attach to a name

Freshness

Machine-readable dates, and whether they are honest

Content

Depth, lists and tables, heading hierarchy, internal links

Each check carries a weight that sums to 100, and the weights do not move between runs. Two audits a month apart are comparable, which is the whole point of a score.

Backlinks, meta and duplicates too

Referring domains come from a full backlink index rather than an estimate, and the duplicate titles, thin pages, orphans and redirect chains come from comparing all your pages against each other — findings no single-page checker can produce. They are reported beside the score and deliberately kept out of it.

[04] THE CONTENT LOOP

From a blindspot to a published page, without leaving the tool.

This is the part most tools stop short of. A blindspot becomes a brief, the brief opens in an editor with the checks running live beside your draft, and the finished piece publishes into WordPress as a draft for you to approve. Then the next cycle tells you whether it moved anything — including when it didn't.

Blindspot

A prompt where rivals are named and you are not.

→

Brief

The questions to answer, the entities to name, the shape to use.

→

Draft

Write it in the editor, with the checks live beside you.

→

Before / after

What the rewrite gained — and what it broke.

→

Publish

Into WordPress as a draft, for you to approve.

BIH2≡⌗🔗Draft · saved 2m ago

Does an ATS actually reject your resume?

Mostly, no. An applicant tracking system parses your CV and ranks it — it does not usually bin it outright. What sinks a resume is a layout the parser cannot read…

Selected · Expand this into a short list

Live checks

✓Answer in the first paragraph
✓Question-shaped H2s
✓A list or table
✓Named entities from the brief
○Heading hierarchy
○Published date in the markup

72/100 content score

Content score

47→78+31

+433 words · content factors only, not a page score

Now passing

+Question-shaped headings
+Direct answer within the first 200 characters
+A scannable list or table
+The four entities the brief asked for

No longer passing

−Clear heading hierarchy — two H2s became H3s

Keywords

addedfree resume builder90,500

addedai resume builder33,100

keptats resume checker12,100

lostresume score checker720

Google search volume — not AI prompt volume, which nobody can measure.

The red list is the point. A rewrite that gains four checks and quietly breaks one is a trade, and you should make it deliberately before you publish rather than discover it three weeks later. We show the losses at the same size as the wins.

[05] SOURCES & TRAFFIC

Where the answers come from, and what reaches your site.

An AI answer is assembled from sources, and most of them are not your website. Knowing that a third of your category's citations come from forum threads changes where the work should go — and it is not something a rank tracker can tell you.

What AI reads before it answers you

Citation sources across your prompts, last 30 days.

Forums & UGC34%

Review sites26%

Documentation18%

News13%

Corporate pages9%

A third of what gets cited is a forum thread you do not control. That is a finding, not a fix — and it changes where the effort should go.

Referring domains, against the field

Root domains from a full backlink index — a count, not an estimate.

rival-a.com73,377

rival-b.io34,664

rival-c.com32,014

rival-d.com12,394

yourbrand.com121

265× behind the median. Stated as a fact, not a verdict — it tells you the size of the gap, not whether closing it is the best use of this quarter.

Organic clicks vs AI referralsOrganicAI referrals

Twelve months, from your Search Console and GA4.

Two scales, because 402 and 3,680 do not share an axis honestly. AI referrals grew 4.6× and are still 11% of organic. Both of those matter.

[06] COMPARISON

How CoreCited compares, with the dates attached.

The headline price in this category is rarely the price of the product people think they are buying. Below is what each entry plan actually includes, read off each vendor's own pricing page rather than a roundup — and the comparison pages behind it list what those tools do better than us, because a comparison nobody believes converts nobody.

Tool	Entry	What that entry price actually buys	Free tier
CoreCited	$69	40 prompts across 6 surfaces — ChatGPT, Claude, Perplexity, Gemini, and Google's AI Overviews and AI Mode. No engine is sold separately and none is held back for Enterprise.	20 checks, no card
Profound	$99	$99 tracks ChatGPT ONLY, with 50 prompts. Multi-engine tracking starts at the $399 Growth plan — so the real entry price for the product most people think they're buying is 4× the headline.	A free one-off AEO report. No free tracking tier; the paid plans have a 14-day trial.
Otterly.AI	$29	$29 Lite gets you 15 prompts. The API and MCP server only begin at the $189 Standard tier.	Free trial only. There is no free tier, despite what several blog roundups claim.
Rank Prompt	$39	$39/mo billed annually ($49 monthly) for 150 credits and 10 brands. Credits meter everything: 10 per AI-generated article, 10 per brand audit, 5 per Lighthouse audit.	7-day trial with 50 credits. No free tier.
Semrush AI Visibility	$99	$99/mo PER DOMAIN (annual), for 25 tracked prompts and 1 domain. Extra prompts are $60/mo per 50; extra domains are $99 each; extra users $45/mo.	7-day trial.
Ahrefs Brand Radar	$398	$398/mo for selected platforms or $699/mo for all — and it's an ADD-ON that requires an Ahrefs base plan (~$129/mo), so realistically ~$527–$828/mo all-in.	No free tier.
Peec AI	on request	Four paid tiers. We are not quoting a number: the widely-repeated €89 figure failed verification against their own page, and we won't publish a price we haven't read ourselves.	No free plan and no trial stated on their pricing page. (Blog claims of a 'free starter plan' were refuted.)

Every figure read from each vendor's own pricing page on 15 July 2026, not from a roundup. Prices in this category change often — if you find one of these out of date, tell us and we'll fix it. Full comparisons, including what each tool does better than us →

[07] LIMITS

What we don't claim.

Four things this product cannot do. They are gaps in what is measurable today, not gaps in our roadmap, and every vendor in this category has them — some just don't mention it. You would rather find out here than after the invoice.

Prompt volume

Nobody can tell you how many people asked an AI a given question. It would take a licensed consumer panel, and anyone quoting you a number is modelling, not measuring.

AI Overview click traffic

No API exposes it. GA4 folds those sessions into organic, so any tool showing you a separate figure has estimated it.

Copilot citations

No public API. We would rather track six surfaces properly than list a seventh we cannot read.

Any guarantee of being mentioned

We measure what the engines say and tell you what to fix. Nobody controls the output of a model they don't own.

[08] PRICING

One price, all six surfaces, no add-ons.

Engines are never sold separately here and nothing is held back for a custom-priced tier. The only thing that changes between plans is how many prompts and brands you track.

Free

$0/mo

✓1 brand
✓20 checks total
✓Claude + Gemini
✓No card needed

Starter

$69/mo

✓1 brand
✓40 prompts weekly
✓All 6 surfaces
✓Full site audit + tasks

Questions people actually ask.

Traditional SEO tools track Google rankings. We track whether AI engines — ChatGPT, Claude, Perplexity, Gemini — actually mention your brand when buyers ask them questions. Ranking #1 on Google no longer means you appear in the AI answer.

Start with 20 free checks.

Enter your domain and see which of the six surfaces name you today. No card, no call, no trial that quietly starts billing.

Start free — no card Run the free checker