Engines change quietly. Your chart doesn't.
Geoptimizer runs the questions your buyers ask on ChatGPT, Gemini, Claude, and Grok with web search on, reads every answer, and plots one 0–100 AI Visibility Score per engine. It marks the day a model changed, so a drop has a cause.
Free · no signup · first scan in about 30 seconds
Why a line moves
A score that only updates when you remember to press a button is a photograph. This is a chart.
Daily rotation
Paid plans re-check every tracked prompt at least weekly on a daily rotation, so the dashboard moves every day. The free plan runs on demand.
Model changes, marked
Engines swap models quietly. The trend line marks the day it happened, so a drop has a cause instead of a guess.
A 7-day window, not one sample
AI answers are nondeterministic. Headline numbers are a 7-day rolling window with a confidence band; single on-demand scans are labeled as snapshots.
Where the traces ended last week
A production scan, not a mock-up. Negative deltas are left in.
- Mention rate
- 92%−3.0
- Answers naming the brand
- Citation rate
- 3%+3.0
- Answers citing the domain
- Prominence
- 0.97
- 1.0 = always mentioned first
- Sentiment
- 0.64
- 1.0 = always promoted
Engines name Notion in 92% of answers but cite notion.so in 3%. That gap is what the score is for.
One formula, on paper
Every number on the chart comes from this. It is versioned, and it really computes.
EngineScore = 100 × ( 0.35 · MentionRate + 0.25 · CitationRate + 0.20 · Prominence + 0.20 · Sentiment ) VisibilityScore = mean(EngineScore) // 7-day rolling window with a confidence band. // Share of Voice uses an open denominator: // every brand the AI names counts.
Worked on the real scan
| 0.35 · Mention 0.92 | 0.322 | named in 92% |
| 0.25 · Citation 0.03 | 0.0075 | linked in 3% |
| 0.20 · Prominence 0.97 | 0.194 | usually first |
| 0.20 · Sentiment 0.64 | 0.128 | mostly promoted |
| 100 × 0.6515 | ≈ 65 | notion.so · Aug 24, 2026 |
We query the official APIs with web search on. API answers approximate what the consumer apps show; the limitation is published on the methodology page.
“Which all-in-one workspace app has the best AI features right now?”
Verbatim excerpts. Cuts are ellipses. The tick is where the brand is first named.
- ChatGPTgpt-5.4-mini
“If you mean “best AI features” in an all-in-one workspace app as of now, my pick is ClickUp. Why: ClickUp currently has the broadest AI surface area…”
first mentionnever named - Geminigemini-3.1-flash-lite
“…1. Best for Knowledge Management & Flexibility: Notion AI. Notion remains the top recommendation for those who prioritize documentation, wikis…”
first mentionnamed 12% in - Claudeclaude-haiku-4.5
“…ClickUp Brain² stands out with particularly impressive AI capabilities… Notion AI is deeply integrated into the workspace: Notion Agent handles complex multi-step tasks…”
first mentionnamed 40% in - Grokgrok-4.3
“Notion AI currently leads among all-in-one workspace apps (docs, databases, tasks/projects, wikis/knowledge bases, and more) for its AI capabilities…”
first mentionopens with it
What feeds the chart, and what fixes it
Prompts decide what gets asked. The site audit explains why engines skip you.
Five questions your buyers type into AI
- What is the best CRM for a small sales team in 2026?
- Which CRM integrates with Slack and Gmail without an admin?
- Is Acme CRM worth it compared with HubSpot Starter?
- What tools do startups use to keep customer notes in one place?
- Cheapest CRM with a free plan for 3 users
Drafted from your own site. Accept the ones that fit, add your own, pin the priorities.
Two engines almost never name Acme
- ChatGPT1 of 5
- Gemini4 of 5
- Claude1 of 5
- Grok4 of 5
Every prompt runs on all four engines with web search on.
robots.txt blocks their search crawlers
OAI-SearchBot and Claude-SearchBot are disallowed, so ChatGPT and Claude cannot read acme.com.
Fix: remove the two Disallow rules.
- warningAnswer-first intro · 158 words, target 40–80
- warningllms.txt · missing an H1 title
- passJSON-LD · and 8 more checks
Fix one, and the chart marks the day.
Free GEO optimization tools
Four free tools for generative engine optimization (GEO), each answering one question in seconds: are you visible in AI answers, can the engines read your site, and what should buyers' prompts look like. No signup, no email wall — every tool shows its full result immediately.
- 01
Instant AI Visibility Check
Runs 3 buyer-intent prompts live on ChatGPT, Gemini, Claude, and Grok and scores whether they mention and cite your brand. Around 30 seconds, 3 checks a day.
Open tool - 02
GEO Site Audit
Audits any page across 5 categories — AI crawler access, llms.txt, rendering, structured data, content shape — with a fix for every failed check. Instant.
Open tool - 03
AI Crawler Access Checker
Reads your robots.txt and llms.txt, then live-fetches your page as the retrieval bots to catch CDN blocks robots.txt can't see. Blocked crawler = engine that can't cite you.
Open tool - 04
Prompt Ideas Generator
Turns your domain into 20 realistic buyer-intent prompts, tagged by funnel stage — the question set worth tracking your AI visibility against.
Open tool
Three plans, one chart
No per-engine add-ons. Annual billing is two months free. The free plan never expires.
Free
$0
- 3 AI-suggested prompts
- All 4 AI engines
- On-demand scans
- 3 spot-checks + 1 full sweep/mo
- Your first visibility report free
Starter
$19/mo
- 10 tracked prompts
- All 4 AI engines
- Daily-updating dashboard
- 6 spot-checks + 1 full sweep/mo
- 2 Premium + 1 Ultra check/mo (flagship models)
- Weekly PDF reports
Pro
Most popular$49/mo
- 30 tracked prompts
- All 4 AI engines
- Daily-updating dashboard
- 8 spot-checks + 1 full sweep/mo
- 3 Premium + 1 Ultra check/mo (flagship models)
- Weekly PDF reports + full history
Daily rotation re-checks each tracked prompt at least weekly. Starter includes flagship-model checks at the entry price. Full comparison
Before you start
What is Geoptimizer and what does it do?
Geoptimizer (geoptimizer.io) is a Generative Engine Optimization platform. It runs your buyers' real questions on ChatGPT, Gemini, Claude, and Grok with web search enabled, measures how often your brand is mentioned and cited in the answers, and turns that into one transparent 0–100 AI Visibility Score, plus prioritized recommendations to improve it. Scans run on demand and a spot-check finishes in about 30 seconds.
What is Generative Engine Optimization (GEO)?
GEO (generative engine optimization) is the practice of making your brand appear, and be cited, in the answers AI assistants like ChatGPT, Gemini, Claude, and Grok generate. You'll also see it written as GEO optimization, generative AI optimization, or geoptimization; they all mean the same practice. It's distinct from geographic or location-based optimization, and it differs from classic SEO in that AI answers have no position one: visibility is measured as the share of runs where you're mentioned, not a rank.
What is an AI Visibility Score?
A 0–100 composite measuring how present, prominent, cited, and positively framed your brand is in AI-generated answers. It combines four components per engine: mention rate (35%), citation rate (25%), prominence (20%), and sentiment (20%). Engines are then averaged into one score.
Which AI engines does Geoptimizer track?
ChatGPT, Google Gemini, Claude, and Grok — all four in every plan, including the free tier. Grok coverage matters because its answers draw on X (Twitter) search, a citation source the other engines don't surface.
How is this different from Otterly or other GEO tools?
Three ways: scans run on demand in about 30 seconds instead of waiting for a daily refresh; all four engines are included in every tier instead of sold as add-ons; and the full composite scoring formula (weights, windowing, versioning) is published on our methodology page.
Is Geoptimizer worth it? How much does it cost?
We'd rather you verify that than take our word for it. The free plan (3 AI-suggested tracked prompts, all 4 engines, on-demand checks) and four ungated free tools never expire, so you can watch your own numbers before paying anything. Paid plans start at $19/month with premium- and ultra-model checks included; Pro at $49/month tracks a 30-prompt pool refreshed daily, while other tools commonly sell individual engines as paid add-ons. No annual lock-in, and the scoring formula is public, so you can audit exactly what you're paying for.
Why does my score change between scans?
Because AI answers are nondeterministic: ask the same question twice and you can get two different answers. That's why the headline score is a 7-day rolling window with a confidence band, not a single sample. Single on-demand scans are clearly labeled as snapshots.
Do API answers match what users see in the apps?
Closely, but not exactly, and we won't pretend otherwise. We query official APIs with web search enabled: the same retrieval systems, identical methodology across engines. The limitation is published openly on our methodology page instead of being buried.
Start your chart.
The first reading takes about 30 seconds. From then on, every model change and every fix gets a mark on the paper.
Start your chartAll four engines · free plan never expires