<aside> 📖
When Hume came to us on October 6th with 7 weeks until Black Friday, most agencies would have panicked or played it safe with surface-level tests. We took a different approach.
</aside>

Buybox experiment for /pages/pdp3
The redesigned buy boxes went live in early November. From November 24 through January 31 (69 days), we tracked two phases of impact: the BFCM surge and the post-holiday sustained lift. Together, they generated nearly $12M in incremental revenue from three page redesigns.
| Product | CVR Increase | Additional Orders | Incremental Revenue |
|---|---|---|---|
| Hume Pod | +16.4% | 11,142 | $2,175,613 |
| Hume Band | +19.73% | 3,492 | $803,166 |
| Total | — | 14,634 | $2,978,779 ✅ |
| Metric | Projected Result |
|---|---|
| Additional Orders | 42,390 |
| Incremental Revenue | $8,728,290 🔥 |
| Avg Revenue per Order | ~$205.90 |
| Period | Duration | Incremental Orders | Incremental Revenue |
|---|---|---|---|
| BFCM Peak | 14 days | 14,634 | $2,978,779 |
| Post-BFCM | 55 days | 42,390 | $8,728,290 |
| Combined Total | 69 days | 57,024 | $11,707,069 🚀 |
We started with depth, not speed. Week one was spent analyzing nearly 2,000 Pod customer surveys, almost 1,000 Band surveys, and over 1,600 live user insights, extracting the exact problems, emotions, and friction points that were costing Hume conversions:
From those insights, we didn't build 15 tests. We identified the highest-leverage experiments: recreating buy box sections (including product hero galleries) for their conversion pages. We rebuilt these sections around credibility (science-backed accuracy claims), clarity (showing what's really changing in your body), and proof of progress (data that validates effort).