Once you’ve verified a winning template in GemX, the experiment phase is over, but the optimization process is not.
This guide walks you through what to do next to ensure you:
-
Apply the winner correctly
-
Avoid data disruption
-
Lock in revenue gains
-
Continue optimizing with confidence
Step 1: Double-check the Winner Before Applying
Before you make any changes live, you need to confirm:
-
The winner has higher conversion rate or revenue per visitor
-
Statistical confidence is strong enough
-
The test ran long enough (no early stop bias)
-
No tracking issues during the experiment
Open the Performance detail panel in GemX for an overall comparison between two testing templates.
If you’re unsure how to interpret results, review:
Tip: Do not apply a winner based only on a short-term spike.
Step 2: Apply the Winning Template
Once verified, you can apply the winner to 100% of the traffic.
To apply the winning template, follow this guidance:
-
Go to GemX Dashboard
-
Open the completed experiment
-
Click Make Winner
-
Select the winning variant
-
Confirm the action
After confirmation:
-
All traffic will be directed to the winning template
-
The experiment will stop collecting data
-
The losing variant will no longer be shown
Learn more: If you need detailed steps, see: How to Apply the Winning Template in GemX
Step 3: Verify Live Behavior on Your Store
After applying the winner, here is the checklist you should do:
-
Open your storefront in an incognito window
-
Clear cache
-
Test on both desktop and mobile
-
Go through the funnel (i.e.: Product → Cart → Checkout)
Make sure:
-
The correct template is loading
-
No broken layout or missing sections
-
Tracking events still fire properly
This is especially important if you were running both Template Testing for page-level tests and Multipage Testing for full-funnel experiments.
Step 4: Document the Insight (Not Just the Result)
A winning template is not just a “better design”, it validates a hypothesis.
Try to ask and define:
-
What exactly changed?
-
Why did it improve performance?
-
Was it messaging, layout, CTA placement, trust elements?
These questions help you turn your test result into a reusable insight.
For example, instead of writing "Variant B won with +8% CR"
You can write:
Simplifying the hero headline and moving the CTA above the fold increased the add-to-cart rate among mobile users.
This allows you to build a repeatable testing framework.
Step 5: Monitor Post-Test Performance
Applying the winner does not guarantee permanent up lift.
It’s a must for Shopify store owners to monitor the key metrics such as conversion rate, revenue per visitor, add-to-cart rate, and also funnel drop-off (if your test is a Multipage Testing one).
Within GemX, you can use...
...to watch performance for at least 1–2 weeks after rollout.
If performance drops during this time period:
-
Check traffic sources
-
Check seasonality
-
Check if behavior has changed
Please note that sometimes, external factors influence post-test results.
Step 6: Decide What to Test Next
Remember: Optimization is iterative. After applying a winner, ask:
-
What bottleneck remains?
-
Which metric is still underperforming?
-
Where is the biggest drop-off?
You can take advantage of GemX's built-in analytics, such as:
-
Page Analytics to identify weak sections
-
Journey Analysis to identify funnel leaks
-
Order Analytics to understand revenue impact
Then, it’s time to create a new hypothesis and launch the next experiment.
This is how you build a testing culture, not just run isolated tests.
Common Mistakes to Avoid
1. Apply the winner too early
Stop the test before statistical confidence leads to false winners.
2. Forget to verify tracking
Always ensure analytics continue working after applying the template.
3. Ignore device differences
Sometimes desktop wins but mobile underperforms. Review device-level data.
4. Not save your experiment learnings
Without documentation, you’ll repeat the same tests later.
Related Articles
FAQs
