Skip to main content
How Can We Help?

Declaring a Winner

Posted
Updated
ByEmily Gomez
Once your experiment has collected enough data and you’re confident in the results, you can declare a winning Variant.
Declaring a Variant as the winner does more than record the outcome of your experiment. Bluvia applies the winning Variant’s page or content changes to the original Control, making the winning experience your new live version.
Because this action changes your original content, review your results and winning Variant carefully before confirming your selection.

When Should You Declare a Winner?

A Variant shouldn’t be declared the winner simply because it currently has the highest conversion rate.
Before making a decision, consider:
  • The number of visitors who participated.
  • The number of conversions collected.
  • The conversion rate of the Control and Variants.
  • Your experiment’s statistical confidence.
  • How consistently the Variant has performed.
  • How long the experiment has been running.
  • Whether the results support your original hypothesis.
Bluvia uses a 95% confidence level by default, although you can adjust the confidence level for your experiment.
Reaching your selected confidence level can provide stronger evidence for your decision, but it should still be considered alongside the rest of your experiment data.

Review the Winning Variant

Before declaring a Variant the winner, review the experience one final time.
Remember that the winning Variant’s content will be applied to the Control after you confirm your selection.
Before proceeding, ask:
  • Did the Variant improve the Conversion Goal you were testing?
  • Is the improvement meaningful enough to implement?
  • Does the result align with your original hypothesis?
  • Were there any unusual traffic patterns during the experiment?
  • Does the Variant look and function as expected?
  • Are you comfortable making this version the new live experience?
Taking a final look helps ensure you’re ready to make the tested changes permanent.

Declaring Your Winner

Once you’re satisfied with the results, select the appropriate Variant as the winner from your experiment report and confirm your selection.
When a Variant is declared the winner:
  1. Bluvia records the Variant as the winner of the experiment.
  2. The winning Variant’s page or content is applied to the original Control.
  3. The Control is updated to reflect the winning experience.
The winning changes are now part of your live WordPress content rather than remaining only within the experiment.

What Happens to the Control?

The Control is the original content used as the baseline for your experiment.
When you declare a Variant as the winner, Bluvia updates that original content with the content from the winning Variant.
For example, if you’re testing two versions of a landing page and Variant A is declared the winner, the changes made to Variant A are applied to the original landing page.
This allows you to move from experimentation to implementation without manually recreating the winning changes.

What If the Control Wins?

Your original Control can also remain the winning experience.
If the Control consistently outperforms your Variants, there’s no need to apply Variant content to the original page or content. You can keep the existing experience and use what you learned to inform your next experiment.
This doesn’t mean the experiment failed.
You’ve learned that the proposed change didn’t improve the Conversion Goal you were measuring, which can help guide your next optimization decision.

What If There Is No Clear Winner?

You don’t have to declare a Variant as the winner for every experiment.
If your Control and Variants perform similarly or your results remain inconclusive, avoid applying a Variant simply to complete the test.
An inconclusive result may indicate that:
  • More data is needed.
  • The tested change had little impact.
  • The Variants were too similar.
  • The selected Conversion Goal wasn’t strongly affected.
  • A different hypothesis may be worth testing.
It’s better to acknowledge an inconclusive experiment than to make permanent changes that aren’t supported by the data.

Before You Confirm

Declaring a Variant as the winner changes the original content, so complete one final review before confirming.
Make sure:
  • You’ve reviewed your sample size.
  • You’ve reviewed your conversion data.
  • You’ve considered your selected statistical confidence level.
  • The Variant has performed consistently.
  • You’ve previewed the winning Variant.
  • The Variant works correctly across the devices that matter to your website.
  • You’re ready for the Variant’s changes to replace the original Control content.
Once everything looks correct, you can confidently declare the Variant as your winner.

What’s Next?

After applying a winning Variant, use what you learned to identify your next optimization opportunity.
A successful result might lead you to:
  • Test another element on the same page.
  • Apply similar changes elsewhere on your website.
  • Develop a follow-up hypothesis.
  • Test another variation of the winning idea.
  • Explore another part of the visitor journey.
A winning experiment isn’t the end of optimization. It gives you new information you can use to make your next experiment even stronger.

Related Articles

Continue with: