Compare assets
Compare assets shows different prototypes to different participants while keeping the same study structure. Each arm can use its own design, start screen, end screen, and success URL mapping.
When to use it
Section titled “When to use it”- A/B test two Figma (or other) prototypes on identical tasks.
- Serve a prototype based on a URL or sample attribute, not only random groups.
- Keep one study, one distribution link, and segmented results by variant.
Preference tests that place designs side by side on a page are a different pattern. Compare assets is for assigning one prototype experience per participant (or per configured rule).
Drivers
Section titled “Drivers”| Driver | When to use |
|---|---|
| Randomization | Balanced experiment arms from randomization groups |
| Condition | Deterministic assignment from URL, participant, or other condition variables |
Page vs section compare
Section titled “Page vs section compare”| Scope | Behavior |
|---|---|
| Page compare | Rules apply on that page. Task mapping stays shared for the page. |
| Section compare | Rules apply to pages in the section that use the source asset. Supports richer task mapping policy. |
Variant task mapping
Section titled “Variant task mapping”For each variant rule you can map interactive task fields:
| Field | Meaning |
|---|---|
| Start | Where the participant begins in the prototype |
| End | Completion / end screen for the task |
| Success URL | URL used to detect success when applicable |
On section compare, mapping policy can be:
| Policy | Behavior |
|---|---|
| Shared start | One start mapping for all pages using the source asset |
| Per page start | Different start mapping by page |
| Shared end / success URL | One mapping for all tasks |
| Per task end / success URL | Different mappings by task block |
Page-level compare keeps mapping shared for that page.
Fallback
Section titled “Fallback”If the driver value is missing or matches no rule, configure fallback behavior:
- Show nothing for that asset
- Fall back to the source prototype
- Fall back to a specific prototype
How to set it up
Section titled “How to set it up”- Add the prototypes your arms need.
- On a page or section, enable Compare assets for the source asset.
- Choose driver: randomization or condition.
- Select the variable (group or condition source).
- Map each variable value to a prototype.
- Configure start, end, and success URL mapping (shared or granular).
- Set fallback behavior.
- Preview each arm, then confirm with Live test takes.
Example
Section titled “Example”Randomization variable ab_group with values A and B:
- Rule A → Prototype Checkout v1 (start Home, success URL
/order/confirm) - Rule B → Prototype Checkout v2 (start Home, success URL
/order/confirm) - Same goal-based tasks and success criteria for both arms
Limitations
Section titled “Limitations”- Compare assets is for usability prototype assignment, not for swapping arbitrary survey media everywhere.
- Changing prototypes mid-study affects comparability.
- Mixed section structures and conflicting page/section configs can confuse which rule wins. Keep one clear host (page or section) per experiment.
- Keep task wording identical across arms when the prototype is the only intended difference.
- Name rules with the same labels as your randomization groups.
- Analyze by the driver variable in Results Studio.
Common questions
What is the difference between page and section compare assets?
Page compare applies to one page. Section compare applies across pages that share a source asset in that section, with optional shared or per-task mapping.
What drivers can select a prototype?
A randomization group variable, or a condition variable such as a URL or participant value.
Can start screens differ by variant?
Yes. Configure variant task mapping for start, end, and success URL as shared or more granular scopes.