Ali Demirbaş - Lab
TRGet in touch

Home & Landing

Welcome Takeover Screen on Page Load: A/B Test Idea

A prominent welcome overlay can focus attention on one message or offer, but it also adds a step before the page.

  • Home & landing
  • Continuation rate

Control vs Variant

Absent on one side, present on the other.

Changed
Presence
Page
Home & landing
Difference
added

Hypothesis

Because only the overlay arm has a close action, closing it cannot be a shared success event. Define the same meaningful interaction with the underlying page for both arms.

How to run this test

Primary metric

Continuation rate

Measure the same predefined interaction with the page below per assigned visitor; do not compare overlay dismissal with page load.

Other metrics

  • Page exit rateCompare exits among visitors assigned to the opening overlay.
  • Offer acceptance rateReport acceptance among eligible assigned visitors in the variant; the control has no equivalent offer.
  • Time to reach main contentMeasure time from page load to first meaningful interaction with the main content consistently in both arms.
  • Repeat visit rateCompare all eligible assigned visitors over the same follow-up period.

What to check before and during the test

  • OverlayAdd one dismissible opening overlay that covers a predefined share of the screen.
  • MessageKeep one message or offer fixed and preserve the page content below.
  • DismissalProvide a visible, usable close control.
  • TimingUse the same opening trigger throughout the test; delayed display is a separate test.
  • Visitor historyDefine eligibility and review new and returning visitors separately.

Setup mistakes to avoid

  1. 1Do not hide the close control or force visitors to accept the offer.
  2. 2Do not change overlay size and content together.
  3. 3Do not configure the overlay to prevent search bots from accessing the main content.
  4. 4On mobile, keep the close target large enough to tap.
  5. 5Do not show the overlay again in the same session after a visitor closes it.