CapstonAI is an AI visibility platform that helps brands track, audit, and optimize their presence in AI-generated answers. It monitors how your brand is mentioned across ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews, then provides actionable fixes to improve your share of voice in AI search results.

How does CapstonAI track AI visibility?

CapstonAI sends structured brand queries to multiple large language models (LLMs) on a recurring basis and records whether your brand is mentioned, cited, or recommended. It aggregates these results into a GEO score, a share-of-voice metric, and a citation report.

Which AI models does CapstonAI monitor?

CapstonAI currently monitors ChatGPT (OpenAI), Claude (Anthropic), Gemini (Google), Perplexity, and Google AI Overviews. Coverage is continuously expanded as new AI search surfaces emerge.

CapstonAI offers a free WordPress plugin that provides basic AI visibility scanning for your site. For full AI visibility tracking, audit reports, automated content fixes, and multi-brand monitoring, the Core Plan is available at $99/month.

How is CapstonAI different from Semrush or Ahrefs?

Semrush and Ahrefs are built for traditional search engine optimization. CapstonAI is purpose-built for the AI search era: it tracks your brand presence inside AI-generated answers from ChatGPT, Claude, Gemini, and Perplexity.

GEO Attribution Model 2026: How To Credit AI Engines In GA4, Mixpanel, Heap, Salesforce

GEO Attribution Model 2026: How to Credit AI Engines in GA4, Mixpanel, Heap, Salesforce

Attribution is where most GEO programs die — not because GEO doesn’t work, but because traffic from ChatGPT, Perplexity, Claude, and Google AI Overviews lands as ‘direct’ or ‘unassigned’ in GA4 by default, making it invisible to finance. Without attribution, no budget. The CapstonAI Q1 2026 cohort that built proper GEO attribution before launch saw 3.2x faster budget approval and 47% higher year-1 ROI than cohorts that bolted attribution on later. Below: complete attribution setup for GA4, Mixpanel, Heap, Salesforce, plus UTM templates, custom dimensions, and the 8-step build.

TL;DR: Build GEO attribution in 8 steps: (1) audit current AI referrer behavior in GA4, (2) create custom dimension ‘ai_source’, (3) deploy UTM templates for any AI-clickable links, (4) tag self-reported survey at signup, (5) join custom dimension to lead/account in CRM, (6) build conversion path report including AI sources, (7) credit assists not just last-click, (8) reconcile weekly with GEO citation data.

Free CapstonAI scan → Pricing

The 8-step playbook with formulas

Step 1: Audit current AI referrer behavior in GA4. Filter Acquisition > Traffic Source by referrer host. Look for: chat.openai.com, chatgpt.com, perplexity.ai, claude.ai, gemini.google.com, copilot.microsoft.com. Most show up as ‘direct’ or ‘referral’ but bucketed wrong. This is your zero state.
Step 2: Create custom dimension ‘ai_source’ in GA4 (event-scoped). GA4 Admin > Custom Definitions > Custom Dimensions > Create. Name: ai_source. Scope: event. Event parameter: ai_source. Then in GTM, add a tag that fires on page_view, sets ai_source = derived from referrer (chatgpt | perplexity | claude | gemini | copilot | google_ai_overview | other_ai | none).
Step 3: Deploy UTM templates for any AI-clickable link. When you can control the link (footer of email, LinkedIn bio, etc.), use: ?utm_source=chatgpt&utm_medium=ai_referral&utm_campaign=geo_org. For AI-discoverable site content, you can’t control inbound UTMs, so rely on referrer + custom dimension.
Step 4: Add self-reported attribution at signup. Add a ‘How did you hear about us?’ field at signup with options including: ChatGPT, Perplexity, Claude, Google AI Overview, other AI assistant. CapstonAI cohort sees 38-52% of GEO-attributed signups self-identify here, making it the most reliable single signal.
Step 5: Join ai_source dimension to lead/account in CRM (Salesforce, HubSpot). Pass ai_source as a hidden form field on signup. Create field on Lead + Contact + Account in Salesforce/HubSpot. Map it through the funnel so deal-level reports show GEO contribution. Critical for CFO reports.
Step 6: Build conversion path report (assisted + last-touch). GA4 Advertising > Attribution > Conversion paths. Filter by ai_source. You’ll see GEO appears in 22-41% of conversion paths even when not last-click — this is the assist value. Credit it in your report.
Step 7: Switch attribution model from last-click to data-driven. Last-click undervalues GEO (which often introduces brand). Data-driven attribution (GA4 default since 2023) credits GEO assists. Mixpanel: use multi-touch model. Heap: use cohort-based attribution.
Step 8: Reconcile attribution data weekly with citation panel data. If your panel shows you’re cited 47% of the time but ai_source dimension shows 0.5% of traffic, attribution is broken. Cross-check weekly. Cohort target: panel citation rate within 3-5x of ai_source traffic share — wider gap = attribution gap.

Concrete benchmarks (CapstonAI Q1 2026 cohort, 86 customers)

Anonymized cohort medians by industry. Use as honest baseline — not stretch targets.

AI source	Referrer host	Default GA4 bucket	Recommended UTM (when controllable)	Custom dim value
ChatGPT	chatgpt.com / chat.openai.com	Referral or Direct	utm_source=chatgpt&utm_medium=ai_referral	chatgpt
Perplexity	perplexity.ai	Referral	utm_source=perplexity&utm_medium=ai_referral	perplexity
Claude	claude.ai	Referral or Direct	utm_source=claude&utm_medium=ai_referral	claude
Gemini	gemini.google.com	Direct (often)	utm_source=gemini&utm_medium=ai_referral	gemini
Copilot	copilot.microsoft.com	Referral	utm_source=copilot&utm_medium=ai_referral	copilot
Google AI Overview	google.com (with AIO)	Organic Search	(not controllable; detect via GSC)	google_ai_overview
Other AI assistants	various	Referral or Direct	utm_source=other_ai&utm_medium=ai_referral	other_ai

Common errors when measuring GEO attribution

Over-attributing to last-click direct. 60% of GEO traffic appears as ‘direct’ without proper setup. Don’t credit it to direct — fix the attribution. Otherwise GEO looks like 0% and dies.
Ignoring assist conversions. GEO often introduces brand 2-4 sessions before conversion. Last-click model misses this entirely. Use data-driven or position-based attribution.
Weekly reporting fatigue from over-detailed dashboards. CFO doesn’t read 30-tab dashboards. 1-page weekly with 5 numbers (sessions, signups, ai_source share, blended CAC, GEO CAC) wins.
Not joining ai_source to deal-level data in CRM. Without CRM join, you can’t credit GEO with closed-won revenue. Set up the field on Lead + Contact + Account before launch.
Trusting GA4 referrer alone. Referrer alone misses 30-50% of AI traffic (privacy headers strip referrer often). Always layer self-reported survey on top.

FAQ — GEO attribution

Why does so much AI traffic appear as ‘direct’?

AI engines often strip referrer for privacy, especially mobile + desktop apps (vs. web). Self-reported survey at signup is the most reliable single signal — it consistently catches 38-52% of GEO traffic that would otherwise be invisible.

Can I attribute revenue (not just signups) to GEO?

Yes — once ai_source flows from form to Lead to Opportunity to Closed-Won in CRM. Cohort median: 18% of closed-won revenue attributable to GEO at the Lead source level by month 6.

How does this work for SaaS with self-serve trial-to-paid?

Same model. ai_source dimension flows from signup event into your billing system (Stripe metadata field, Chargebee custom field, etc.). MRR by ai_source becomes your headline GEO ROI metric.

Tools and related reading

Ready to measure GEO attribution properly?

Free CapstonAI scan →

Last updated: May 2026. Sources: CapstonAI Q1 2026 cohort (86 customers, 24 800 LLM responses analyzed), GA4 official documentation, Mixpanel + Heap attribution docs, Salesforce + HubSpot CRM field documentation, Looker Studio connector docs, Search Engine Land x CapstonAI analysis.