- What Happens When You Make a Winner
- Before You Apply a Winner
- Choose the Right Winner (Control vs Variant)
- How to Apply the Winning Variation (Step-by-Step)
- What Happens After Applying the Winner (By Page Type)
- Important Notes & Limitations
- Quick Checklist After Applying the Winner
- FAQs about Make Winner feature in GemX
Once your experiment has collected enough data, the next step is to decide which version should go live.
The Make winner action in GemX allows you to:
-
End the experiment
-
Select the better-performing version (Control or Variant)
-
Apply the winning design to your live store
This is a critical step in your testing workflow. Applying a winner will update your storefront, so it’s important to understand how the process works before confirming.
What Happens When You Make a Winner
When you apply a winning variation in GemX, the system does not simply switch from one template to another.
Instead, GemX keeps your original template (Control) and applies the winning design by updating it behind the scenes.
How it works
After you confirm a winner, GemX will:
-
End the experiment
-
Declare the selected version as the winner
-
Create a backup version of your original template
-
Copy the design (JSON structure) from the winning variant into the original template
-
Apply the updated template to your live store
This approach ensures that:
-
Your existing template structure is preserved
-
Changes are applied safely without breaking template assignments
-
You can revert manually if needed, using backup versions
Because this action directly affects your storefront, you should always review your experiment results carefully before making a final decision.
Before You Apply a Winner
Before applying a winning variation, take a moment to review your experiment results and confirm that the outcome is reliable.
1. Review your experiment results
In the Analytics view, focus on the following:
-
Primary metric: Identify which version performs better based on your main goal (e.g., Conversion rate, Revenue).
-
Supporting metrics: Compare additional metrics such as Average Order Value (AOV) or Revenue per Visitor (RPV) to ensure there are no negative trade-offs.
-
Probability to win: Check the confidence level for each variant. A higher probability indicates more reliable results.
This step helps ensure that the selected winner aligns with your overall business objectives, not just a single metric.
2. Confirm your decision
Before ending the experiment:
-
Make sure the performance difference between versions is meaningful
-
Avoid selecting a winner too early when the data is still limited
-
Consider whether the results are consistent over time
If you are unsure, continue running the experiment until you reach a clearer outcome.
Learn more: For more guidance, refer to How to Read Your Experiment Results.
Choose the Right Winner (Control vs Variant)
When you click Make winner, you’ll need to choose between the Control (A) and the Variant (B). Each option leads to a different outcome.
Understanding this distinction helps you avoid unintended changes to your live store.
Option 1: Select Control (A)
If you choose the Control:
-
The experiment ends immediately
-
No changes are made to your live store
-
Your current template remains active
Use this option when:
-
The Variant does not outperform the Control
-
The results are inconclusive
-
You decide not to apply any changes after reviewing the data
This is the safest option if you are not confident in the test results.
Option 2: Select Variant (B)
If you choose the Variant:
-
The process to apply the winning template begins
-
You will be guided through additional confirmation steps
-
The winning design will be applied to your existing template after confirmation
Use this option when:
-
The Variant clearly outperforms the Control
-
The results are statistically reliable
-
You are ready to update your live store
In most cases, this is the option used to implement improvements identified through your experiment.
How to Apply the Winning Variation (Step-by-Step)
Follow these steps to safely apply the winning variation and end your experiment.
Step 1: Open your experiment
-
Go to the Experiments tab in your GemX dashboard
-
Select the experiment you want to finalize
-
Open the Analytics view
You can click on the experiment name or click the chart icon to access its analytics
Step 2: Click “Make winner”
In the summary section at the top of the report, click Make winner.
This action starts the process of ending your experiment and selecting a final version.
Step 3: Select your winner
Choose one of the following:
-
Control (A): ends the experiment with no changes
-
Variant (B): continues the process to apply the winning design
If you select Variant (B), GemX will guide you through the next steps.
GemX will open the Process tab and begins the flow to apply template B as the winner.
Step 4: Confirm and stop the experiment
In this step, click Stop experiment and review the warning message carefully.
The message indicates that the winning design will replace the content of your current template.
Then, you can click the Continue button to confirm.
Once confirmed, the experiment will be finalized and cannot be resumed.
Step 5: Wait for processing
After confirmation, GemX will automatically:
-
Create a backup of your current template
-
Copy the design from the winning variant into the original template
-
Apply the updated template to your live store
-
Mark the experiment as ended
No manual action is required during this step.
Step 6: Review the success screen
After processing is complete, you will see:
-
The experiment marked as Ended
-
The winning variant clearly identified
Depending on your page type, you may see additional actions or recommendations.
What Happens After Applying the Winner (By Page Type)
After you confirm a winning variation, what happens next depends on the type of page used in your experiment.
GemX handles each page type differently to match how Shopify templates work. Review the scenarios below to understand what to expect.
Homepage
If your experiment is running on a homepage template, you can set the winning version as the default homepage template.
Once applied, the winning design becomes live across your storefront homepage
In most cases, no additional action is required after confirmation.
Product and Collection pages
If your experiment is running on product or collection templates, you will have two options:
-
Apply the winning template as the default template (affects all products/collections)
-
Assign the winning template to selected items only
In case the original template is already set as default:
-
The system may overwrite the default template directly
-
A backup is created before any changes are applied
Important notes:
• Sections with manually selected products may not transfer correctly
• You may need to open the builder and reassign products in affected sections
Landing pages and Blog posts
If your experiment is running on landing pages or blog posts, please note that Shopify does not support a “default template” for these page types.
As a result, it’s required for you to assign the winning template manually to specific pages.
After applying the winner:
-
Select which pages should use the updated template
-
Confirm the assignment to make the changes live
If template types conflict
If the Control and Variant use different template types (for example, GemPages vs Shopify native):
-
GemX will block automatic template replacement to prevent errors
-
The experiment will still be marked as ended
To apply your preferred version, you can manually assign templates in Shopify admin or GemPages after the experiment ends
This ensures your store structure remains valid and avoids unintended issues.
Important Notes & Limitations
Before applying a winning variation, be aware of the following limitations and behaviors. These are based on how Shopify templates and GemX processing work.
Template overwrite and backup
Applying a winner may overwrite your current template, especially if it is set as a default (e.g., product, collection, or homepage).
GemX automatically creates a backup version of the original template before making changes
This allows you to manually restore the previous version if needed.
Product section limitations
Sections that use manually selected products may not transfer correctly when applying the winning design.
This is common for product blocks, bundles, or custom sections
If this happens:
-
Open the template in the builder
-
Reassign the correct products in affected sections
Shopify template limitations
Shopify enforces limits on JSON structure and number of sections. Therefore, large or complex templates may fail to apply or require adjustments
In some cases:
-
You may need to simplify the template
-
Or manually fix sections after applying the winner
Cache delay
Please note that your changes to your storefront may not appear immediately, as Shopify caching can delay updates by 1–5 minutes.
If you do not see the updated version:
-
Wait a few minutes
-
Perform a hard refresh (Cmd + Shift + R or Ctrl + F5)
Quick Checklist After Applying the Winner
After applying the winning variation, review your storefront to ensure everything is working as expected.
Use the checklist below to validate the update:
1. Verify the page on your storefront
-
Open the updated page on your live store
-
Check both desktop and mobile views
Make sure the layout and content display correctly across devices.
2. Test key functionality
Confirm that important actions still work properly:
-
Add to Cart
-
Navigation (menus, links, buttons)
-
Images, banners, and media loading
This helps ensure the applied design does not affect the user experience.
3. Check product-specific content (if applicable)
If you applied the winner to a product template:
-
Review sections that include manually selected products
-
Confirm that the correct products are displayed
If anything looks incorrect:
-
Open the template in the builder
-
Reassign products in affected sections
4. Monitor performance after applying
After the update goes live, you need to monitor key metrics (such as conversion rate, revenue, and add-to-cart rate) over the next 24–48 hours.
If you notice unexpected drops, review the template setup and section configurations.
FAQs about Make Winner feature in GemX
