Sebastian Käufer
OPEN FOR NEW PROJECTS Book a call

Work / EV-01 / Cro · mobile product page

Mobile add-to-cart from 1.95% to 5.08% in 28 days.

One product page, three changes, and an honest read of what the numbers can and cannot say.

Client
Nicebeam, red light therapy, US
Service
CRO retainer, diagnosis to build
Windows
27 May – 23 Jun vs 24 Jun – 21 Jul 2026
Stack
Shopify theme, GA4, Clarity
EV-01 · ADD-TO-CART RATE, 28 DAYS EACHBEFORE ○ → AFTER ●
Changed page, item-scoped1.95 → 5.08% · +161%
Changed page, cross-network traffic only3.66 → 10.31% · +182%
Untouched page A, same traffic cut5.9 → 6.3%
Untouched page B, same traffic cut10.2 → 12.2% · noise
GA4, ITEM-SCOPED · CHANGED PAGE 36/1,848 → 72/1,418 SESSIONS · Z ≈ 5.0
EV-01b · STORE-WIDE +50% Session-to-order rate 0.69% → 1.04%, completed checkouts 102 → 149, on traffic that moved −2.5%.
WHERE IT CAME FROM Traffic mix explains about 5% of the store-wide change, about 95% happened within the same sources.
EV-01c · SCROLL REACH BEFOREIPHONE 390 × 844 · 14 DAYS
5%
99.2
10%
60.5
20%
32.0
30%
22.1
50%
15.4
100%
1.6
% OF VISITORS REACHING EACH PAGE DEPTH · ~40% GONE BEFORE 10% · #1 TAP: POPUP CLOSE BUTTON
EV-01d · FIRST SCREENSCHEMATIC
POPUP
ADD TO CART
BEFORE · BUY BLOCK BELOW FOLD
− 1 +
ADD TO CART
AFTER · QTY + CART IN ONE ROW
DASHED LINE = FOLD · POPUP MOVED FROM ON-LOAD TO 40% SCROLL DEPTH

01 — FINDING

The page with the most traffic converted worst. On an iPhone the add-to-cart sat below the fold, about 40% of visitors left before scrolling a tenth of the page, and the most-tapped element was the close button of a newsletter popup.

02 — WHAT I CHANGED

  1. Quantity and add-to-cart in one row, above the fold
  2. Details open by default, spec icons moved down
  3. Popup from on-load to 40% scroll depth, store-wide

03 — HOW IT WAS MEASURED

Shipped as a reversible template, read every 14 days against the 28 days before. The two untouched product pages served as the comparison.

04 — WHAT THIS DOES NOT PROVE

Strong and directional. Not a randomized test.

  • No traffic split: the page changed for everyone at once.
  • The popup change went live two days later and is part of the effect.
  • GA4 lost data for 9–12 July, inside the after window.