Analytics Overview: Experiences & Experiments

Understand Jurni analytics for individual experiences and experiments, including sessions, conversion rates, revenue, attribution, and how to compare performance responsibly.

Jurni analytics should help you answer a marketing question—not just give you another dashboard to watch.

Start with what you are trying to learn:

How is this page performing?

Use experience analytics when you want to understand the traffic and outcomes associated with one Jurni experience.

Which treatment performed better?

Use experiment results when traffic was deliberately split across variants and you want to compare the treatments.

That distinction matters. A page can receive direct traffic, campaign traffic, or experiment traffic. An experiment view is specifically about the traffic assigned through that test.

The metrics marketers use most

Metric	The question it helps answer
Sessions	How many measured visits are in this view?
Orders	How many attributed purchases came from those sessions?
Revenue	How much attributed revenue came from those sessions?
CVR	What share of sessions became orders?
RPS	How much revenue did each session generate on average?
AOV	How much revenue did the average attributed order contain?
CTR	What share of sessions included a tracked click?
Add-to-Cart Rate	What share of sessions reached add to cart?
Checkout Rate	What share of sessions reached checkout?

Use the Metrics glossary when you need the exact definition or formula.

Start with the business outcome, then diagnose the journey

If the campaign goal is purchase conversion, start with CVR or the relevant revenue metric. Then use earlier steps to understand where behavior changed.

For example:

Low CTR can mean the page is not making the next action compelling or obvious.

CTR is healthy but Add-to-Cart Rate is weak can point to product, offer, or purchase-choice friction.

Add-to-Cart Rate improves but CVR does not means more people reached cart without a matching increase in orders. Look further down the journey before declaring the page better.

CVR is flat but RPS improves can mean the treatment influenced revenue per visit through order value or product mix.

These are diagnostic patterns, not automatic explanations. The numbers tell you what changed, not necessarily why.

Use the same context when comparing performance

Before comparing two periods or views, check:

• date range
• campaign / traffic source
• audience
• offer
• product availability
• market / currency
• major page changes
• consent / tracking changes

A “better” week can simply contain different traffic or a stronger offer.

Experience analytics versus experiment results

Experience analytics

Use this when you want to understand the performance of a page itself across the traffic reaching it.

Good questions:

• How much traffic did this experience receive?
• What is its CVR / RPS?
• Where do shoppers progress or drop off?
• Did performance change after a page update?

Be careful when comparing two separate experiences that received different traffic. A higher CVR does not automatically mean one design caused the difference.

Experiment results

Use this when you intentionally routed comparable traffic across variants through the experiment’s Smart Link.

Good questions:

• Which variant leads on the metric we chose for this hypothesis?
• Did the treatment change earlier journey behavior?
• What did this test teach us?

For the complete marketer workflow, use Run your first A/B test.

Pick your primary metric before you chase the dashboard

A simple decision guide:

• CVR — purchase conversion is the main question.
• RPS — you care about the combined revenue effect of conversion and basket value.
• AOV — the treatment is specifically intended to influence order value among purchasers.
• CTR / ATC / Checkout Rate — the hypothesis is focused on an earlier step, or you are using these as diagnostics.

If the treatment changes subscription mix, bundles, quantities, or upsells, revenue metrics may tell a different story from CVR alone.

Why Jurni can disagree with GA4, Meta, or another analytics platform

Different platforms can use different:

• session definitions
• identity / cookie rules
• attribution windows
• consent handling
• channel attribution
• timezone and date boundaries
• event deduplication

That means totals do not need to match exactly to be useful.

When investigating a large difference, align the date range and traffic first, then check the measurement path. Use Why analytics platforms differ and Attribution overview.

Turn analytics into a next action

Use this prompt when you want Jurni AI to help organize what you see:

Here is the page context and Jurni analytics:

TRAFFIC / AUDIENCE
[context]

PAGE / OFFER CHANGES
[context]

METRICS
[paste relevant values]

Do not claim a cause from correlation.
Separate:
1. OBSERVATIONS — what changed
2. POSSIBLE FRICTION POINTS — hypotheses worth investigating
3. WHAT ELSE I SHOULD CHECK
4. 3 TESTABLE NEXT HYPOTHESES

Prioritize customer decisions over cosmetic changes.

Then use Review a page before testing or Decide what to test next to turn the observation into a clean experiment.

Continue

• Experience analytics
• Conversion-step performance
• Reading experiment results
• Filtering analytics
• Metrics glossary