Applies to beta version Price Test, URL Redirect Test, Theme Test, Template Test, Shipping Test , Visual Editor Test
Please make sure you're using beta version before referring to this article.
The new beta version uses a standard flow for all test types, that a visitor will need to pass the following steps to enter your experiment.
1. Traffic allocation
When a visitor arrives at your store, ABConvert first decides if they should enter the test at all. You control this with the "Traffic allocation" percentage (1-100%).
100% = All eligible visitors see the test (recommended)
Lower percentages = Useful for cautious testing or limited inventory
For example, if you set traffic allocation to 50%, half of your visitors will enter the test and half won't. This decision is sticky - once a visitor is included or excluded, they stay that way.
2. Audience filter (optional)
After traffic allocation, you can apply additional filters to narrow down who sees your test. Audience filters only apply to new visitors who haven't been assigned yet.
Basic Filters:
Countries - Target specific geographic regions
Traffic Sources - Include/exclude visitors from specific channels
Available Channels
Paid Search
Paid Social
Organic Search
Organic Social
Email
SMS
Affiliate
Referral
Direct
Available Platforms
Google
Facebook
Instagram
TikTok
Twitter/X
YouTube
Devices - Target desktop or mobile users
Visitor Type - Choose between new visitors or returning visitors
Please make sure to enable ABConvert Visitor Label embedded apps at least 2 weeks before you use this filter in any test.
Advanced filters: (Click "Advanced filters" to expand)
Referral Domain - Filter by the referring website domain
Landing Page URL - Target visitors arriving at specific pages
URL Parameters (UTM) - Target visitors with specific UTM parameters
Cookie Values - Target based on existing cookie data
Each advance filters can include complex logic setting such as Landing Page URL starts with /products and contains /A or /B or /C by using the or and and settings.
How filters work together
A visitor must pass ALL enabled filter groups to be able to enter group split.
Example configuration:
- Country: United States OR Canada
- Device: Mobile
- Traffic source: Facebook OR Instagram
Result: Only mobile visitors from US/Canada arriving from Facebook/Instagram will enter.
3. Targeting rules (optional)
You can create custom rules to assign specific visitors to particular test groups in targeting:
Add Assignment Rules - Click "+ Add rule" to create targeting conditions
Set Conditions - Choose from:
Country
Traffic Source
Referral Domain
Landing Page
Device
UTM parameter
Cookie
Select Actions - Assign to group / Random split / Exclude
Assign to group - Direct matching visitors to specific group (control / variant)
Random split - Enter next traffic split step to a random group
Exclude - Exclude from the test and will not be tracked in test analytics
Example Rule: "IF Device is Desktop → Assign to Control"
Default Assignment: Visitors who don't match any rules are randomly split according to your traffic allocation settings.
4. Random group assignment
If a visitor passes all filters and doesn't match any force assign rules, they're randomly assigned to a test group based on your traffic split percentages.
How it works: ABConvert creates a deterministic hash and maps it to your group percentages. This ensures the same visitor always sees the same variant across sessions.
Example: For a 50/50 split between Group 0 and Group 1, visitors with hash 0-49 see Group 0, and 50-99 see Group 1.
Traffic flow diagram
Visitor arrives
↓
Traffic allocation (e.g., 80%)
↓ (passes)
Audience targeting (optional filters)
↓ (passes OR no filters set)
Force assign rules (optional priority rules)
↓ (no match OR assign to random)
Random group assignment (based on percentages)
↓
Visitor sees variant
FAQ
Why traffic distribution may not be exactly 50/50
Even with a 50/50 split setting, your analytics might show 52/48 or similar imbalances. This is normal and expected with the following reasons:
1. Audience filters reduce the sample
When you add audience targeting or force assign rules, you're filtering the traffic pool. Fewer visitors means more variance.
Example: With 100,000 visitors and no filters, you'll see very close to 50/50. With 1,000 visitors after filtering, you might see 48/52 - both are statistically normal.
2. Short-term randomness
Traffic splitting is like flipping a coin. In the short term, you might get 6 heads and 4 tails. Over thousands of flips, it evens out.
What to do: Let your test run longer. As sample size increases, distribution naturally balances.
What if I am running multiple tests simultaneously
Each test independently assigns visitors. The same visitor can be in different groups across different tests.
Example scenario:
- Price Test A: Visitor assigned to Group 0 (Original)
- Theme Test B: Same visitor assigned to Group 1 (Variant)
- Shipping Test C: Same visitor assigned to Group 2
This is by design - each test's assignment is isolated and deterministic based on its own experiment ID.
How to keep visitors in the same group across all tests
Currently it's not possible to make sure all visitors in the same group across all tests. We're planning to launch grouping tests function that will unlock this use case soon!
Can I change traffic allocation after launching
Yes. You can adjust traffic allocation percentage at any time. However, visitors who were already excluded will stay excluded (sticky decision).
Example: If you start at 50% allocation and increase to 80%, the additional 30% will come from previously excluded visitors.
Can I edit audience filters after launching
Yes, but changes only affect new visitors. Visitors already assigned will keep their original assignment.
Best practice: Test your filters thoroughly before launching to avoid losing attribution data.
What happens when a visitor doesn't match audience filters
They're permanently excluded from the test. Excluded visitors won't see any test variants and won't appear in analytics for this test.