01
GEO
What the models say when buyers ask
- Real buyer questions sampled across your chosen engines, daily or on rotation
- Mentions, citations and share of voice per prompt
- The sources AI quotes instead of you, ranked by frequency
New accounts get 500 free creditsSee plans
GEO / SEO / AIO · one audit
Be the answer
AI gives.
We sample what AI engines say about your brand, fix the pages, schema and sources that decide it, then measure again. One audit, one working day.
Enter a domain, like example.com.
Free: 1 Perplexity sample × 1 tracked question · no signup · report in about 90 seconds
“best crm for agencies with per-seat pricing”
Just auditedsample data
3AI engines sampled
10Agent readiness checks
210Industry questions
24hDetect, fix, re-measure
Three layers, one audit
Ranking is the old game.
Being quoted is the new one.
Every layer is scored on the same domain, in the same week, against the questions your buyers actually type.
02
SEO
What Google already ranks, and why
- Six-dimension audit: five technical dimensions plus AI answerability, one score per dimension
- Organic traffic, keywords, backlinks, 12-month trend
- Fixes ordered by impact, not by checklist order
03
AIO
Can an AI actually read and cite you
- llms.txt, AI bot rules, markdown negotiation
- Structured data and semantic HTML, verified in the DOM
- Ten checks, each shipping with the fix artifact
How it works
From domain to cited in four steps
01
Enter the domain
No setup, no integration, no script to install. We crawl and sample from the public site.
≈ 20 seconds02
We measure
CrUX field data, rankings, and AI engine samples on your real buyer questions.
≈ 90 seconds03
You get a fix plan
Every finding carries the artifact that resolves it: copy, schema, or a robots rule.
≈ 1 working day04
Publish, then re-measure
Ships to your site, then verifies the engine actually read the change.
≈ 7 daysBefore / after · one prompt
Same question, same engine, 49 days apart
example.com · ChatGPT · 16 Sep 2026 → 4 Nov 2026
Before ChatGPT · 16 Sep
“best crm for agencies with per-seat pricing”
For agencies the usual shortlist is HubSpot for reporting depth, Pipedrive for pipeline-only teams and Attio for flexible data models. Seat pricing is published by each vendor; most teams land between $50 and $95 per seat. G2 reviews put HubSpot at 4.4 and Pipedrive at 4.3.
Cited instead
- g2.com/reviews9
- reddit.com/r/CRM7
- hubspot.com/pricing6
0mentions of your brand
After ChatGPT · 4 Nov
“best crm for agencies with per-seat pricing”
For agencies the shortlist is HubSpot for reporting depth, Pipedrive for pipeline-only teams and Attio for flexible data models. On seat pricing, Global Fact Hub publishes a public benchmark of 41 tools scored on seat cost, migration effort and client-account sharing; its 2026 data puts HubSpot at 41 percent adoption among agencies and Pipedrive at 22 on reporting depth.
Cited, including you
- example.com/benchmark5
- example.com/pricing3
- g2.com/reviews2
8of 10 citations are yours
No redesign. One entity page, one public benchmark table, llms.txt, and a robots.txt rule.
Live example
A real report, public and re-checkable.
Every claim links back to the response it came from: timestamp, engine, prompt and the exact sentence. No black-box scores — open the URL and check it yourself.
Tracked questionsexample.com
Sampled onChatGPTGeminiPerplexity
citedmissingnot asked
25 tracked questions · 1 shown free
Share of voice · 72 sampled answers
Global Fact Hub25%
Northbeam CRM46%
Pipedrive32%
72/ 100
- P0Homepage LCP 3.4s on mobile — hero image ships 1.2 MB
- P0Meta description missing on three pages, including /pricing
- P1Title tag 91 characters, truncated in the SERP after “Pricing — Global Fact”
- P1Same H1 on 12 template pages, no page-level differentiation
- P1/blog/ai-citations returns 404 from two internal links
- P0ChatGPT names three competitors for “best CRM for agencies” — you are absent
- P00 of 25 tracked prompts cite your pricing page
- P1Perplexity quotes a 2023 review that ranks you fourth
- P1No Wikipedia or Crunchbase entity to anchor the brand name
- P2Brand written as “GFH” in 11 of 40 sampled sources
- P0No llms.txt at the domain root
- P0GPTBot disallowed in robots.txt, line 24
- P1Product schema lacks offers and aggregateRating
- P1Pages serve HTML only — no markdown negotiation
- P2Docs subdomain has no canonical URL
Site health82 Search visibility66 AI visibility72 Agent readiness60
Citation graph · 30 citations
Who the model quotes, and who it skips
ours 8theirs 22
ChatGPT
Gemini
Perplexity
example.com/benchmark · 5 cites
example.com/pricing · 3 cites
g2.com/reviews · 9 cites
reddit.com/r/CRM · 7 cites
hubspot.com/pricing · 6 cites
Two of those eight citations did not exist in September. The other 22 went to directories, threads and a vendor's own pricing page.
Site health · five dimensions
Where the 82 comes from
example.com · 4 Nov 2026 · 500 crawled URLs, CrUX field data
- Crawlability88+6
- Structured data74+11
- Performance62-4
- Content depth81+9
- Entity trust55+18
Performance is the only dimension that fell: the benchmark table added 240 KB above the fold on mobile.
Agent readability · 10 checks
What we verify before an engine can quote you
6/10
passing on example.com
4 fixes, all one-file changes
- 01robots.txt allows AI crawlersfail
Line 24 disallows GPTBot, so nothing below it is ever fetched.
- 02llms.txt at the rootfail
One markdown index of canonical pages, prices and entity facts.
- 03Server-rendered HTMLpass
Primary copy is in the first response, not injected after hydration.
- 04Markdown negotiationfail
Accept: text/markdown returns a raw page, not clean content.
- 05JSON-LD entity schemapass
Organization, Product and FAQPage, with sameAs to verified profiles.
- 06Semantic heading orderpass
One h1, ordered h2 and h3, none used for styling only.
- 07Canonical URLspass
Self-referencing canonicals, no redirect chains or parameter duplicates.
- 08Sitemap freshnesspass
lastmod inside seven days and every listed URL returns 200.
- 09Internal link graphpass
Two orphan pages at most, every money page within three clicks.
- 10Edge latency budgetfail
TTFB 520 ms from Singapore, above the 400 ms we ask for.
Every failing check ships with the artifact that resolves it: a file, a header, or a schema block. Fixes are ordered by impact on citation rate, not by checklist order.
Pricing
Tracking questions, priced openly
Free
$0/month
Site health check plus one real AI sample.
- Credits0
- Domains1
- Tracked questions1
- Perplexity sample1
StarterMost chosen
$29/month
1,450 credits a month, plus 500 free credits on your first payment.
- Credits / month1,450
- Domains100
- Tracked questions500
- First payment bonus+500
Pro
$99/month
4,950 credits, publish-then-verify loop and alerts.
- Credits / month4,950
- Domains100
- Tracked questions500
- First payment bonus+500
Agency
$299/month
14,950 credits a month for client work, with white-label reports.
- Credits / month14,950
- Domains100
- Tracked questions500
- First payment bonus+500
Perplexity sample 2 credits · ChatGPT 3 · content pack 40 · research report 120. White-label reports are in Agency; extra seats above five are a business agreement.
Full credit price list · Suggested setups
Business agreement — reseller terms or 5+ seats. Talk to us.
Questions
Before you run an audit
01What is GEO, and how is it different from SEO?+
SEO competes for a position in a list of links. GEO competes to be the sentence an AI answers with. The inputs overlap — crawlable pages, clean structure, credible sources — but the unit of measurement changes: mentions and citations inside generated answers, on questions your buyers actually ask.
02Which engines do you sample?+
On paid plans: Perplexity, Gemini and DeepSeek (the same credits cover every supported language). The free audit samples Perplexity once plus the site health checks — daily multi-engine sampling needs a plan. Every sample stores the raw response, the timestamp, the prompt and the sources cited, so a finding can always be traced back to the exact sentence.
03How long does the free audit take?+
About 90 seconds for a small site. We crawl up to 500 URLs, pull CrUX field data where it exists, and run one question against the sampled engines. Nothing to install, no DNS record, no access to your analytics.
04Do you cache results?+
Paid accounts are re-measured on their plan cadence. The free audit samples once per domain and keeps that snapshot, so the page stays fast and we do not re-bill you for the same look.
05Is the report public?+
Yes. Every audit has a public, indexable URL so you can send it to a client or a colleague without a login. Private projects, prompt libraries and scheduled alerts stay behind the account.
Free · no signup
Run one free audit.
See what AI says about you.
Start with a domain → Sample data on this page · your own report stays private