UI Elements
Opening Links in a New Tab During a Flow: A/B Test
Opening an external link in a new tab can leave the original form or checkout available while visitors read another page. It also changes how they return and may be confusing, especially on mobile. Preserve entered data and clearly indicate the new-tab behavior for the defined external links.
A/B test scenario. This page is a test plan, not the result of a real experiment.
- UI elements
- Completion of the original flow
Control vs Variant
What the interface does on its own is the variable.
- Changed
- Behavior
- Page
- UI elements
- Difference
- changed
Hypothesis
If the predefined external links open in a new tab, the form or checkout stays open behind them and visitors can pick up where they left off after reading. Completion of the original flow may improve, though switching tabs may confuse some visitors on mobile and weaken the effect.
Test plan
When to run this test
A funnel drop right after clicks on help, policy or external-resource links, or session recordings of visitors who cannot find their way back to the form or checkout step, are good reasons to try this test. Visitors who open the link should then return to the original tab and finish the flow. If returning visitors lose the data they entered, fix data preservation first; do not use a new tab to patch that problem or to keep visitors from leaving the site.
How to measure success
Completion of the original flow
Compare completed form or checkout actions among all eligible assigned visitors after they use one of the predefined external links.
Also track
- External-link click-through rateMeasure clicks on the tested external links among all eligible assigned visitors.
- Return to original flowCompare assigned visitors who return to the original form or checkout and continue within the same window.
- Flow data-loss rateTrack cases where entered form data or checkout state is lost in either arm.
How to set it up
- TargetOpen the same predefined links in the current tab or a new tab.
- ScopeFix the eligible help, policy or external-resource links before launch.
- ExpectationKeep the link label; notify visitors that the new-tab version opens a new tab.
- ReturnTrack resumed activity and completed actions.
- DeviceReview actual tab switching and return behavior on mobile.
What to watch when reading the result
- 1A rise in external-link click-through or in return to the original flow does not justify shipping by itself. Completion of the original flow is the decision metric; also check that the flow data-loss rate does not climb in either arm. A gap that comes from lost data means you have not measured tab behavior.
- 2The result covers only the help, policy and external-resource links you fixed before launch, so do not treat it as a case for opening every link on the site in a new tab. Look at each device on its own, since switching tabs and coming back work differently on mobile.
- 3Only some visitors use the tested links, so the comparison can stay thin and the difference unclear. Find the visitors per arm you need for the effect you care about with the sample size calculator, then divide by daily eligible traffic per arm to plan how long the test must run.
Frequently Asked Questions
Which links should be included?
Predefine suitable help, terms or external-resource links leaving a form or checkout. Do not open every site link in a new tab. Preserve wording and destination except for the required behavior notice. Use original-flow completion as the primary outcome.
Should users be told that a new tab opens?
Yes. Add the required notice while preserving descriptive link text. Unannounced behavior can confuse users, especially on mobile. Review actual tab-switching and return behavior.
Why does preserving form data matter?
Preserve entered data and checkout state in both groups. Otherwise the test may measure data loss rather than tab behavior. Track losses and fix persistence rather than using a new tab to conceal the problem.
What should be monitored besides original-flow completion?
Compare return and resumed interaction with the original form or checkout among all eligible assigned visitors within the same window. Measure external clicks only for the tested links.
More tests in UI Elements
All testsDo Links Inside Body Text Get Clicked? A/B Test Idea
UI elementsCompletion of the destination action
The same path offered as an inline link inside a paragraph or as a CTA button below it. See both arms of the experiment, the risks and the tracked KPI.
Adding Shortcuts for Frequent Actions: A/B Test Idea
UI elementsSuccessful task completion rate
An experiment example adding shortcuts such as reorder or reload last search. See the control, the shortcut arm, the trade-offs and the tracked KPI.
Auto-Advancing to the Next Step: Flow A/B Test Idea
UI elementsFlow completion rate
A multi-step flow experiment example: manual continue versus auto-advance once a choice is made. See both arms, the control-and-undo notes and the KPI.