Pega
← All experiments
LossLuhxePDPLX-JAN04

Roller as an add-on

Before
Before
After
AfterAfter
-$422K/yr
Projected annual revenue
-3.2%
Conversion
-2.6%
Rev / visitor

Hypothesis

Repositioning the Roller variant as an add-on upgrade to the Classic bag (instead of a variant selector) will increase AOV and RPV without negatively impacting conversion rate, because framing it as an optional enhancement with clear benefits and a lower price point ($50 vs full bag price) reduces decision complexity and positions the Roller as incremental value rather than a competing choice.

What we learned

**Test Result:** Lost -3% CVR, -3% RPV overall **1. The Add-On Positioning Backfired** • Roller adoption increased modestly (+3.8pp from 59.8% to 63.6%) • But Classic orders dropped -17% (604 to 502) • Overall conversion dropped -3%, killing RPV **Insight:** Framing Roller as an "add-on" made Classic feel incomplete, causing Classic-intenders to abandon rather than convert. **2. More Decision Steps = More Friction** • Control: 1 choice (Classic or Roller?) • Variation: 2 choices (pick base variant, then decide on upgrade) • Result: sequential decisions increased drop-off despite aggressive discounting (72% off messaging) **Insight:** Adding steps, even when framed as "optional upgrades," creates abandonment. Simple variant parity outperformed base + upgrade hierarchy. **3. Traffic Intent Matters More Than Channel** **Winners:** • Organic Social: +18% RPV (low-intent discovery traffic) **Losers:** • Returning Visitors: -13% RPV • SMS: -18% RPV • Email: -9% RPV • Paid Social: -4% RPV **Insight:** The add-on positioning only worked for ultra-low-intent browsers. All high-value, high-intent segments performed worse; you can't optimize for casual browsers at the expense of your best customers. **4. AOV Gains Don't Offset Conversion Losses at This Price Point** • AOV increased +$1.35 per order • But lost 55 orders from the CVR drop • Net impact: -$9,000+ in revenue **Insight:** At $200 AOV, small CVR drops (2-3%) require massive AOV lifts (15-20%) to break even. Conversion always beats AOV optimization.

Want an agency that shows you its losses?

Every agency has losing tests. We're the ones who publish them. Free teardown: we walk your store live and show you the three highest-impact tests we would run first.

$0$500