No Change works as a control group and safety net in AI Optimize and Test optimizations.
The No Change variation shows your original page content to a portion of visitors in AI Optimize and Test optimizations. It acts as a baseline to compare against, and as a failsafe to help protect conversions if new ideas underperform in AI Optimize.
About No Change
The No Change variation is automatically created when you launch a AI Optimize or Test optimization. It mirrors your unmodified, original page content and is treated as a unique variation that can receive traffic and influence results.
No Change can’t be edited or deleted because it ensures accurate measurement and protects performance.
Good to know
Visitors chosen to see one optimization’s No Change may still see variations from another optimization. No Change only acts as the control group for its own optimization.
How it works in Tests
In Test optimizations — the traditional A/B/n type — No Change acts as your control group. Traffic is statically split between variations, including No Change. The percentage of visitors who see No Change stays fixed based on the traffic split you set.
- Acts as a baseline — helps measure how each variation performs compared to your original content
- Supports statistical analysis — allows the system to determine whether variation changes drive meaningful results
How it works in AI Optimize
In AI Optimize, No Change acts as both a control group and a failsafe (i.e., fallback) for the optimization. It helps measure performance and gives the AI a reliable option if other variations don’t perform well.
- Control group — helps measure which versions perform better than the original page
- Failsafe — when all other variations perform worse, more traffic is shifted to No Change to prevent conversion loss
- Dynamic traffic allocation — the percentage of visitors shown No Change changes over time based on performance
Holdout vs. No Change
Both the Holdout group and the No Change variation show your original content — but their purposes and scope are different:
- Holdout group — shows the original content across your entire site; used to measure global optimization impact
- No Change — acts as a control in a specific optimization; used to compare performance of variations in that optimization
Important
Visitors in the Holdout group never see any optimized content making them an unbiased comparison group.