Escaped visitors browsed 2.8× deeper — then bought 2.6× more often.
20 days, 53,784 randomized Instagram visitors. The in-app browser wasn't just hurting checkout — it was killing product discovery: 17.5% of control visitors ever viewed a product, versus 48.7% of escaped visitors.
This was the cleanest split in the portfolio: 27,323 vs 26,461 visitors (a 51/49 coin flip), consistent lift on every single day of the window, near-identical AOV in both arms, and ~100% session identification in both buckets.
The mechanism showed up one step before checkout. Only 17.5% of control visitors ever reached a product page — the webview bounces people before the store gets a chance. Escaped visitors hit product pages 48.7% of the time. Everything downstream (cart, checkout, purchase) inherits that gap.
Checkout conversion: 3.21% escaped vs 1.24% control, a +159% lift at z = 15.4 — about as far from noise as ecommerce data gets.
After the test: The brand moved to a 90/10 rollout after the window (90% escaped, 10% holdout) and is billed on measured performance against the locked 50/50 baseline.
| Arm | Visitors | Orders | CVR | Rev / visitor |
|---|---|---|---|---|
| A · escaped | 27,323 | 876 | 3.21% | $1.54 |
| B · control | 26,461 | 327 | 1.24% | $0.51 |
Brand anonymized. Revenue shown trimmed by the standard outlier rule on both arms — the control arm contained a single $51K wholesale-sized order (median order ≈ $36) that no per-visitor metric should lean on; conversion counts are untouched by trimming.
Run this exact test on your traffic.
One script tag, 60-second install. Randomized 50/50 from the first visitor — in 7–14 days this page is your data.