Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.trodo.ai/docs/llms.txt

Use this file to discover all available pages before exploring further.

​
Overview

Event analytics is the practice of recording specific user actions — events — and using them to understand behavior at scale. Every click, submission, upgrade, and abandonment can be an event. Over time, events form a picture of how your product is actually used, where people get stuck, and what drives retention. Trodo’s event layer is designed for modern AI-native products: a zero-config auto-capture baseline, a flexible SDK for custom instrumentation, and analysis tools built for the non-linear journeys that agentic products create.

​
Events, properties, and identity

Events are discrete actions: signup_completed, agent_run_started, workflow_published, plan_upgraded. Each event has a name, a timestamp, and optional properties — structured data that makes the event filterable and breakdownable. 
Trodo.track('agent_run_completed', {
  workflow_id: 'weekly_report',
  outcome: 'success',
  duration_seconds: 28,
  model: 'gpt-4o',
});
Identity ties events to a person. When you call Trodo.identify('user-42'), every event before and after that call in the session attaches to that user’s profile. Trodo’s CQ Intelligence merge engine handles anonymous → identified transitions and cross-device stitching automatically. User properties (set via Trodo.people.set) are persistent traits on the profile — plan, signup date, company, feature flags — that you can filter and segment on across any report.

​
What gets captured automatically

Enable auto-events on install and Trodo immediately captures a behavioral baseline — no instrumentation required:
EventWhat it captures
page_viewEvery page load and SPA navigation
element_clickAny click with element context
rage_click3+ clicks on the same element in 1s — frustration signal
dead_clickClick on a non-interactive element — broken UI signal
page_scrollNew max scroll depth reached
form_submitHTML form submissions
form_abandonedForm touched but not submitted on page exit
js_errorUncaught JS errors and unhandled promise rejections
page_performanceLCP, CLS, INP, FCP, TTFB — once per load
See the full auto-events catalog for every event and its properties.

​
How it fits into your product

Event analytics answers the human side of the product:
  • Who signed up this week, from where, and how far did they get?
  • Of users who ran an agent for the first time, what share came back within 7 days?
  • Where do users drop out of the onboarding funnel?
  • Which feature correlates most with users upgrading from free to pro?
For the AI side — what the agent actually did, which LLM calls were made, where errors occurred — see Agent Observability. Both layers share user identity, so you can segment and filter across them in one place.

​
Key capabilities

Auto-capture

Page views, clicks, scroll, forms, errors, and performance — zero config baseline from day one.

Custom events

Instrument the actions that matter for your product: runs, upgrades, tool use, completions.

User profiles

Persistent traits that travel with the user across sessions, devices, and time.

Groups

B2B-ready: assign users to companies, workspaces, and teams. Analyze behavior at the account level.

​
Next steps

Quickstart

Install and track your first event in under 5 minutes.

What to track

Plan events and properties for your specific product shape.