Step One Life · US · May 2026
A “Meet the Founder” tile on the PDP lifted conversion for first-time visitors (+5.7%) while actively hurting returning customers and CRM traffic (−44% RPV on SMS). The aggregate hid the mechanism; the segment read turned a no-rollout into a targeting brief.
CVR among new visitors — the tile works for cold audiences
CVR among returning visitors
RPV on SMS traffic — the highest-intent cohort
CVR on direct traffic — brand-aware, not yet loyal
Contents
Step One's founder is a genuine asset — recognisable from social campaigns, with a personal brand story that builds trust. If a "Meet the Founder" tile is added to the PDP, conversion should rise as that familiarity transfers to the product page.
The answer turned out to be: yes for some visitors, at a cost to others — and the segment split is the whole story.
One change: a founder-story tile on the PDP, US storefront, mobile in practice. ~31k visitors across two arms.
| Metric | Change |
|---|---|
| Conversion rate | +2.5% (directional) |
| Add-to-cart rate | +1.6% |
| Average order value | −3.0% |
| Revenue per visitor | −2.1% |
| Revenue | −2.3% |
A directional win on volume metrics, a loss on value metrics — and not remotely significant: at the observed effect size the test would have needed years to reach significance at US traffic levels. The aggregate read alone would have been a shrug.
| Segment | CVR | RPV | Read |
|---|---|---|---|
| New visitors | +5.7% | +1.1% | the tile builds trust with cold audiences |
| Returning visitors | −7.5% | −11.2% | loyal customers don't need the story |
| Direct traffic | +9.7% | +2.9% | brand-aware, not yet loyal |
| SMS traffic | −28.7% | −44.4% | highest-intent CRM cohort, actively hurt |
| Email traffic | −5.0% | −11.4% | warm audience, tile adds friction |
| Paid Social | −7.1% | −5.5% | already saw the founder in the ads |
The hypothesis was right about mechanism and wrong about audience. First-time visitors convert more when the founder story validates the purchase. Visitors who already know the brand — returning customers, SMS and email traffic, people who arrived via the founder-led social creative itself — are given a distraction on the page where they were ready to buy.
No rollout as-is. The directional conversion lift did not compensate for the AOV and RPV decline, and the net revenue outcome was negative.
But the new-visitor signal was strong enough to design the follow-up properly rather than discard the concept: a variant conditionally shown to first-visit sessions only, suppressed for CRM-driven traffic, with placement and prominence tested down. That is the difference between "the test lost" and "the test told us exactly who the feature is for."
Aggregates hide the mechanism. A flat headline concealed a +5.7% and a −7.5% pulling against each other. The segment read is what turned an inconclusive test into a targeting brief.
Brand assets have an audience boundary. The same founder story that converts a cold visitor is clutter to a loyal one — and the most valuable cohort (CRM traffic) was the one most damaged. Rollout on the aggregate would have systematically taxed the best customers.
Source: Shoplift experiment report US_PDP_CVR_260501, Step One US storefront, 1–26 May 2026. All figures traceable to the named experiment ID.