Analytics dashboard

Indie launch analytics and experiments

A launch analytics dashboard that summarizes funnel conversion, source attribution, and experiment results without publishing visitor-level data.

Funnel report

Step-level conversion metrics come from aggregate captured events

Linked funnel
Conversion window
Modeled18%

Warm list opt-in

180 conversions from 1000 visitors.

StageWarm list opt-in
SignalsVisitor activity to conversion activity
Modeled11.9%

Sales page to checkout

74 conversions from 620 visitors.

StageSales page to checkout
SignalsVisitor activity to conversion activity
Modeled68.9%

Checkout start to purchase

51 conversions from 74 visitors.

StageCheckout start to purchase
SignalsVisitor activity to conversion activity

Use these rates directionally until more traffic accumulates.

Event taxonomy

Analytics inputs resolve to stable public-safe event IDs

Audience automation
page viewPrivate fields excluded

Funnel page viewed

Count by route, step, variant, normalized UTM fields, and coarse referrer host after browser-side session idempotency plus server-side bot filtering.

Private data excluded7 fields
opt inPrivate fields excluded

Opt-in created

Count consenting opt-ins by form, segment, and lead magnet.

Private data excluded5 fields
checkout startPrivate fields excluded

Checkout started

Count checkout attempts by product, price, and checkout kind.

Private data excluded4 fields
purchasePrivate fields excluded

Purchase completed

Sum revenue and count paid checkouts from trusted webhook evidence.

Private data excluded5 fields

Source attribution

Source attribution stays aggregate-only

Audience automation
0aggregate source rows
Source window

Raw event rows, visitor keys, full referrers, and raw query strings stay excluded.

No source rows captured yetFuture page views with safe UTM or referrer evidence will appear here as aggregate counts.

Experiment model

Deterministic assignment can be audited before traffic writes exist

Checkout offer
Experimentassignment_ready

Opt-in hero promise test

Hash experiment id plus a caller-provided anonymous assignment key, then map 0-49 to A and 50-99 to B.

VariantsOutcome-first promise 50% / Speed-first promise 50%
SafeguardsSample size

No automated winners

Bumpgrade can summarize assignment counts and decision evidence, but traffic routing waits for sample-size checks and owner confirmation.

Privacy and safety

Metrics are useful without exposing visitor-level data.

Developer details

Aggregate reporting

Conversion and source views summarize patterns without publishing raw event rows.

No raw event feed

IP addresses, user agents, cookies, contact IDs, Stripe IDs, and raw event rows stay out of public data.

Experiment changes are protected

Visitor tracking and experiment decisions open only after privacy review, sample-size checks, retention limits, owner confirmation, and redacted audit evidence.