Open any running or finished test from your dashboard to see how it's performing. The top of the page gives you a quick verdict: which variant is ahead and how confident Shogun is in that result. Underneath, the Insights tab lets you dig into every metric, filter by date, device and traffic source, and see what's driving the numbers.
The test summary
The top of the test page summarises the whole test:
Status and confidence. The status (for example, Running), how long the test has been live, and a badge showing Shogun's current confidence, such as "81% Confidence in positive result".
Test type and goal. The kind of test (for example, Page test) and the goal metric that decides the winner (for example, Clickthrough).
Notes. A place to record your hypothesis or changes.
Variants. Each variant's share of traffic, total sessions, and its result on the goal metric. Challengers also show a Variant Improvement figure, which is the difference compared with the Original.
Below the variants, a confidence meter explains where the test stands in plain language. For example: "Variant B is currently winning. We're 81% confident it will be the overall winner of this test. We are aiming for 95% confidence to decide a winner. We recommend you run this test for a few more days."
You'll also see a preview of each variant, with a Winning label on the variant currently leading.
The summary and the confidence meter always describe the whole test. The filters on the Insights tab change the metric cards below them, but they don't change the confidence level or which variant is winning.
How Shogun decides a winner
Shogun calculates confidence for your test's goal metric. Confidence is Shogun's estimate of how likely the leading variant is to remain the winner once the test is complete. Shogun aims for 95% confidence before it declares a winner.
If a test hasn't reached 95% yet, it usually needs more traffic or more time. Early results can swing, especially on low-traffic stores or with purchase-based goals, so avoid ending a test the moment one variant pulls ahead. When you're ready to end a test, see Publishing Winners.
Insights and Timeline
Below the variant previews, there are two tabs:
Insights shows every metric for the test, with charts and filters. Most of this article covers the Insights tab.
Timeline lists everything that has happened to the test and who did it, such as creating the test, duplicating a template, starting, pausing or ending it. See Editing Active Tests for more on this.
Metrics on the Insights tab update about every minute, and the confidence level and winning variant about every 15 minutes. If your numbers don't match Shopify or GA4, open Why is my data different than Shopify? just above the tabs, or read Why Analytics Look Different in Shogun A/B Testing vs. GA4 or Shopify.
Filtering your results
Use the filters at the top of the Insights tab to narrow what the metric cards show:
Date range. Choose Entire test, Today, Yesterday, Last 7 days or Last 30 days, or pick a custom range on the calendar. Options longer than the test has been running are greyed out.
Devices. All devices, Desktop only, or Mobile only.
Showing. Group the charts by hour, by day, or by week. Grouping by hour is only available for shorter date ranges.
Traffic sources. Direct, Paid social (including Meta Ads, Google Ads and other paid social), Organic social, Agentic (visitors arriving from AI assistants), Email, Referral and Organic search. You can select more than one.
If you've connected Meta Ads or Google Ads, you can also drill into individual campaigns from the Traffic sources filter.
What changes when you filter by traffic source
When you select one or more traffic sources, every card recalculates for that traffic only. You can also see how each source contributes:
Hover over a chart to see, for each variant, how much of that day's result came from each selected source.
Switch a card to a bar chart to see each bar split by traffic source, with a color key above the chart.
Filtering can leave you with a much smaller group of visitors. With fewer visitors, percentages can swing sharply on very few orders, so check the session and order counts before reading too much into a large difference.
Reading the metric cards
Your test's goal metric always appears first, marked with a Goal badge. The other metrics follow in their own cards.
Each card shows:
The overall value across all variants.
Each variant's value, labelled A, B and so on.
An improvement badge showing how the challenger compares with the Original: green with an up arrow when it's ahead, red with a down arrow when it's behind.
A chart of the metric over time. Hover over the chart to see the values for a specific day or period.
Click the ⓘ icon next to a metric's name to see what it measures. For full definitions, see Test Goal Definitions.
Cards with more than one metric
Related metrics share a card. Click the tabs at the top of the card to switch between them:
Profit and Per visitor (profit per visitor)
Add to carts and Add to cart rate
Page views, Sessions and Visitors
Orders, Sales and RPV (revenue per visitor)
Metrics such as Average Order Value and Conversion Rate have their own cards.
Breakdowns
Some cards show more detail below the chart. On a phone, click the second of the two dots under the chart to see it. For example:
A clickthrough goal shows the top clickthrough destinations and top referrers for each variant.
Conversion Rate shows the top products ordered from each variant, with quantities and revenue.
Each list starts with the top entry. Click Show top 5 to expand it.
How profit is calculated
The Profit card takes the revenue from each variant's orders and subtracts two kinds of costs. You can use either or both:
Cost of goods sold (COGS). What each product costs you to make or buy. Shogun takes this from the Cost per item field on each product or variant in Shopify, under Products → [product] → Pricing.
Fulfillment costs and fees. Fixed costs you set once in the Shogun app, under Settings → Set fulfillment costs and fees. They apply to every order across your whole catalog: an average Shipping and fulfillment cost per order, plus Other fees for transaction and processing costs, either as a percentage or a flat fee. They're a quick way to get a realistic profit figure if you haven't set a cost per item on your products.
Labels under the Profit figures show which costs are included, for example, fulfillment costs when you've set them in Settings. If you haven't set fulfillment costs and fees, Shogun uses $0, so profit can look higher than it really is. If you change your fulfillment costs while a test is running, the new values are reflected in that test's analytics.
If some products in the test don't have a cost per item in Shopify, you'll see a red COGS missing label with a count. Click it to see which products are affected, add a cost per item to them in Shopify, then click Check again. Changes can take some minutes to appear.
Card options
Each card has a few controls:
Line chart / Bar chart. Use the two chart icons on the right of the card to switch views.
Pin card. Click the pin icon at the top left of a card to move it up so it sits directly below the goal card. Pinned cards apply to every test you view, whatever the test type, so the metrics you care about most are always near the top. Click the pin again to unpin it. The goal card can't be pinned because it always comes first.
Share image. Click the share icon at the top right of a card to save a picture of it for a report or a message to your team.
On a narrow screen, such as a phone, these options move into the ⋮ menu on each card.
Exporting results
To share the full results, click More actions next to End test at the top of the page, then choose Export to CSV or Export to PDF. CSV is best for your own analysis; PDF gives you a clean snapshot for presentations.
Troubleshooting
If your results don't look right, check the following:
Not enough traffic yet. Small samples can skew early results. Give the test more time, especially if your goal is purchase-based or your store gets less traffic.
Check the filters. A narrow date range, device or traffic source filter can make numbers look unusually low or volatile. Set them back to Entire test, All devices and All traffic sources to see the full picture.
Check your goal. Make sure the goal metric matches what the test is trying to change.
Audience targeting. A narrow audience can limit how many visitors enter the test.
Uneven traffic split. If one variant is getting noticeably more visitors than your chosen split, see Understanding Sample Ratio Mismatch (SRM).
Still stuck? Use Get email support or Start live chat at the bottom of the test page and we'll take a look.