Rheo documentation
User Guide

Customers & analytics

Per-flow analytics — funnels, drop-off, completion by segment, and response summaries across your audience.

What flow analytics answers

After you assign a flow through a channel, the analytics tab on that flow shows how the population behaves:

  • How many people started and completed
  • Where they dropped off in the sequence
  • How long steps took
  • How completion varies by platform, language, app version, experiment variant, or a custom property from your app

For a single user's timeline, use Customers. For visitors, sessions, and retention across the whole app, use Analytics.

Flow analytics overview with headline metric cards, environment selector, and date range

Open flow analytics

  1. Go to Apps → (your app) → Flows and select a flow.
  2. You land on Analytics (not the editor).
  3. Choose environment (test or live) and a date range (presets such as last 30 days or a custom window).
  4. Optionally filter to a specific published version when you want metrics for one pin only. All versions blends data across pins according to how events were recorded.

Some cards compare to the prior period of the same length — helpful after a copy change or new publish.

Headline metrics

Typical cards include:

MetricMeaning
StartsCount of flow_started events in the selected environment, date range, and version. This counts events, not distinct people.
Completion rateflow_completed divided by flow_started over the same filters.
Median time to completeMedian time from the first event in a session to flow_completed, for sessions that finished the flow.
Worst dropoff stepAmong sessions with no flow_completed, each session counts once at the last screen or paywall surface it opened. The tile names the step with the largest share.

Rheo uses completion language rather than generic "conversion" so onboarding goals stay clear.

Flows with a RevenueCat, Superwall, or Stripe node also show Purchases, Paywall conversion, and Revenue. Purchases and revenue are originating webhook gross for that flow. Paywall conversion is surface_presented events with a surface_outcome of purchase_completed, divided by surface_presented. Customer spend, on the customer profile, is the net of originating charges, renewals, and refunds, so it will not match the flow revenue card.

Funnel and drop-off

Each chart reflects one flow (or experiment arms when you compare variants). It is not a separate object you create in Rheo.

  • Funnel — Steps in order with counts between screens. See cumulative throughput through the sequence.
  • Drop-off — Highlights steps where disproportionate abandonment happens. Prioritize builder fixes here.

Step duration

Charts summarize time spent in each step — useful for spotting heavy media, slow network calls, or confusing UI that keeps people stuck.

Completion heatmap

The heatmap shows completion intensity across dimensions Rheo returns for your filters (for example locale or platform). Use it to find segments that lag the overall rate.

Cohort completion over time

Cohort charts plot completion over calendar time, split by:

  • Platform (iOS, Android, …)
  • Locale
  • App version
  • Experiment variant — compare arms while a test is running or after stop
  • Custom property — a key your developers emit from the SDK (for example subscription tier)

When you pick custom property, Rheo selects the first available key automatically so the chart is not empty by accident.

Cohort completion chart split by experiment variant or platform

Responses and field summaries

Where the UI exposes them, response panels summarize answers from input fields and similar controls — validate option wording and defaults without exporting raw events.

How channels and experiments feed these charts

Analytics reflect who actually received the flow through your channel assignments:

  • A flow pin on a channel sends everyone on that channel to one published version.
  • A flow experiment that the channel is serving tags events with the variant each person was bucketed into. Filter cohort charts by experiment variant to compare arms.
  • A code pin or code experiment does not render a flow, so it does not appear in these flow funnels. Code experiment rates use exposures. See Experiments.
  • Test vs live environments are fully separate; always match the environment to the channel you care about.

App product analytics

Flow analytics above are one published flow. Apps → (your app) → Analytics is the app itself. How to read those reports is in Analytics.

The same reads are available from the terminal with rheo analytics product. See Product analytics and the CLI.

Limits

  • You need read access to the app workspace.
  • After large traffic spikes, aggregates may take a short time to catch up.