Mobile App
Does an Exit Intent Pop-Up Work on Mobile? A/B Test
An exit signal suggests that a visitor may be leaving; it does not prove they will. An offer at that moment may encourage them to continue or interrupt someone who would have stayed. Compare the presence of the popup at the same eligible trigger; this does not compare exit intent with a timer.
A/B test scenario. This page is a test plan, not the result of a real experiment.
- Mobile app
- Continuation rate after trigger
Control vs Variant
The moment is the variable, not what appears.
- Changed
- Timing
- Page
- Mobile app
- Difference
- added
Hypothesis
If an offer shown at the moment of an exit signal gives a visitor who is about to leave a reason to stay, that visitor may keep going. If its pull outweighs the cost of interrupting people who would have stayed anyway, that could raise the continuation rate after trigger.
Test plan
When to run this test
When mobile exits cluster on certain pages in analytics, or session recordings show visitors leaving in the middle of the content, an exit-intent offer has a real problem to work on. If the offer works, more triggered visitors go beyond closing the pop-up and move on to the predefined next activity. If exits come from slow or broken pages, fix those first, because an offer will not solve that. Confirm the exit signal works reliably on mobile, keep the pop-up out of the payment flow, and make sure no other pop-up opens in the same session.
How to measure success
Continuation rate after trigger
Record the same trigger in both arms and compare the predefined next activity among eligible assigned visitors; do not count dismissing the pop-up as continuation.
Also track
- Pop-up response rateTrack the share who accept the offer.
- Purchase conversion rate (CR)Check the impact on overall purchases.
- Session time after triggerRead time alongside meaningful activity; a longer session may reflect interruption.
- Newsletter unsubscribe rateIf the offer collects subscriptions, track later unsubscribes over the same follow-up period.
How to set it up
- TriggerUse one predefined exit signal and record it even when no pop-up is shown.
- Mobile eligibilityExclude devices without a reliable signal for this setup.
- OfferKeep the offer, copy and pop-up design fixed.
- AudienceKeep eligibility and any cart-based segmentation fixed.
- FrequencyUse one fixed frequency cap and do not reopen the pop-up in the same session.
What to watch when reading the result
- 1A high pop-up response rate or longer session time after trigger is not success on its own, because an interruption can also stretch a session. The continuation rate after trigger picks the winner; check overall purchases with conversion rate, and if the offer collects subscriptions, compare newsletter unsubscribes across an equal follow-up window.
- 2A result from this exit signal, offer and pop-up design says nothing about a timer-based pop-up. Do not trust a result from a version where the close button was hidden or shrunk, and confirm it stays visible and easy to tap on every screen.
- 3With a small gap, first make sure you compared only visitors who were triggered in both arms and that trigger logging works the same way in each. Also confirm that triggered visitors split across the arms as planned (sample ratio mismatch). If the split holds, use the A/B test significance calculator to see whether the gap in continuation rate after trigger goes beyond chance.
Frequently Asked Questions
Does closing the exit pop-up count as continuation?
No. Use a defined meaningful activity after the common exit trigger. Log that trigger in control too. Compare equivalent triggered visitors rather than counting dismissal or longer session time as success.
Should this exit pop-up run on mobile?
Exclude contexts without a reliable exit signal through predefined eligibility. An unreliable mobile trigger can interrupt visitors who were not leaving.
Do more accepted pop-up offers mean a positive result?
Not alone. Check purchasing and meaningful continuation. For newsletter offers, monitor unsubscribes over equal windows. Additional acceptance can coexist with greater interruption.
More tests in Mobile App
All testsTesting an App Install Banner on Mobile Product Pages
Product pageCross-channel revenue per visitor
An experiment idea that adds an app install banner to mobile product pages and reads revenue per visitor, not installs alone, to judge the trade-off.
Showing First-Run Tooltips in a Mobile App A/B Test
Mobile appPromoted feature adoption rate
Experiment example layering short tooltip bubbles over an app on first use, tracking adoption of the highlighted feature and the risk of cutting the flow.
Testing a Permission Priming Screen Before the Prompt
Mobile appNet permission grant rate
A test idea explaining why a permission is needed before the system dialog appears, with net opt-in rate as the metric and the extra step as the trade-off.