Reading your results

What every number on the Split Tests and Overview tabs means, and when it is worth acting on.

Luria shows you the same numbers it uses to decide. This page reads them with you, field by field, and tells you when a number is worth trusting.

What you'll need

Your impact. Revenue added (an estimate that counts once a test ends with a winner), tests concluded and visitors tested, for this month or the last 30 days.
Your tests. Every test with its type, status, date, visitors, uplift (marked "early" while it runs) and result ("Collecting" until it is decided). Click a row for every number.
The test running now. The original and the new version side by side, each with its conversion rate, and how far along the test is. Under it, a plain line says what Luria still needs, for example how many visitors each version needs before it can call the test.

Open the Classic view fold for the per-test card Luria decides on:

Name and where. What the test changes, the page it runs on and the element.
Verdict. "Ahead" or "Behind" once the chance is past 80% or under 20%, "Too early" in between, then "Winning", "Losing" or "No winner" when the test is decided.
Version table. Visitors, bought, rate, revenue and traffic share for the original and the new version. Visitors counted here are visitors who were actually shown that version.
Chance the new version wins. The probability that the new version is genuinely better than the original, given the counts so far. Above 95% is the promotion bar. Below 5% and the new version is stopped.
The gap in pp. The difference between the two rates in percentage points, as a middle estimate rather than a fact. One percentage point on a 2% conversion rate is a 50% relative change, so read pp carefully.

Traffic stays split evenly until each version alone is past 200 visitors. Allocation is recalculated every hour.

Added by tests. Money the finished tests added. It reads "not yet" until a test is decided.
Tests running and Tests finished. One test runs at a time, biggest first.
Your tests. Each test with its status, visitors, result and a money range. A running test's range is wide on purpose: it is where the result could land, 95% sure.
Before vs after Luria (folded). Your store against its first 4 weeks. "Change vs before Luria · measured" is a real comparison; "not yet measured" means Luria does not have that baseline yet.
Your store numbers (folded). Orders, revenue and visitors who buy, with the day Luria started marked. It will not match Shopify exactly; see Numbers don't match Shopify. Revenue is net of fully refunded orders.

Early numbers swing because the samples are small, not because something is broken.

Under 200 visitors per version the split is held even on purpose, and the chance the new version wins can wander from 20% to 80% on a handful of sales.
A single large order moves tracked revenue much more than it moves the conversion rate.
"Before vs after Luria" stays on "not yet measured" until you have four weeks of data behind the baseline.

Both versions are past 200 visitors. That is the bar Luria promotes on, and it is the right bar for you too.
The chance the new version wins is above 95% or below 5% and holding, rather than crossing back and forth day to day.
The verdict says "Winning". That means the sample bar and the 95% bar were both cleared.
The comparison says "measured". Anything labelled an estimate is a modelled range, not a measurement.

A test shows 0 visitors. The change is live but no visitor has been shown it yet, or serving is off. See No variants serving.
"Before vs after Luria" never leaves "not yet measured". The baseline needs four weeks of data. Low traffic stretches every window.
Conversions look too low. Luria counts attributed orders once each, net of full refunds. See Orders missing and Duplicates.
The chance sits at 50%. The two arms are behaving the same. That is a real result, and Luria will keep the test running until the gap is decidable.