What discount experiments have subscription apps actually run?

The finding: three recurring discount moves

Nearly a third of detected paywall experiments touch a discount or offer.^source[1] Reading the diffs, three moves recur:

Each is an observed before/after, with an inferred reason.^{source[2]source[3]source[4]}

Breakdown: what each team changed

Drops replaced pure urgency ('Last chance') with benefit copy plus a higher reference price ('Previously $199.99') and neutralized the button to 'SAVE 50%.'^source[2] Endel added a second discounted tier (Lifetime $124.99 + Yearly $59.99) with concrete savings anchors, and ran a separate 'BACK TO SCHOOL SALE' surface.^source[3] Peloton swapped a free trial for a time-limited bulk deal ('3 months for the price of 1').^source[4] Walmart anchored on a discounted yearly ('50% off annual at $49 first year then $98/yr') instead of a monthly trial, with a skip-to-trial link.^source[5] Themify made the discount the dominant visual ('50% DISCOUNT' headline) while neutralizing the button to 'Continue.'^source[6]

How to apply it

Steal the structure, then A/B it yourself. The most transferable ideas: (1) pair the discount with a credible higher anchor ('Previously $X') rather than urgency alone;^source[2] (2) offer a second discounted tier so buyers choose between deals, not whether to buy;^source[3] (3) express the discount as concrete value ('3 months for the price of 1', '$49 first year') not just a percentage.^{source[4]source[5]} Because these are detected diffs and not measured outcomes, run each as a real experiment on your own funnel before committing.

Caveats

Every experiment here is a detected before/after UI diff with LLM-inferred rationale — not measured A/B lift.^source[1] The 31.1% counts experiments that touch a discount/offer, not experiments that won. Frame every takeaway as 'observed change + inferred reason,' and treat the named companies as precedent for what to test, never as proof of what works.

Real in-market screens

Screens from real, in-market apps, tracked by Lazyweb Research.

The numbers

Methodology

Universe: 795 detected paywall-CTA experiments across 252 tracked mobile apps, July 2026. Method: keyword identification of discount/offer-touching diffs plus qualitative reading of named before/after changes. Caveat: all rows are detected UI diffs with inferred rationale, not measured A/B lift.

Sources

Source: Lazyweb Research — proprietary analysis of real, in-market app screens. Cite as Lazyweb Research, 2026-07-07.

Explore the evidence