CASE STUDY · 2025 · BEAUTY
8 WEEKS · CRO + EXPERIMENTATION

A conversion lab that pays for itself.

espressOh had the traffic. What it lacked was a system for turning that traffic into bigger, more frequent orders. I built a conversion and blend lab that maps the funnel, tracks basket size, and runs experiments with a clear scoreboard.
CONVERSION & AOV LAB
1

Funnel, basket, experiments

EVERY TEST LOGGED
SHIP-MEASURE-DECIDE

Winners rolled, losers reverted

MADE A DELIBERATE LEVER
AOV

Bundles and thresholds tracked

The problem.

Plenty of people landed on the site and plenty of them left. Conversion and average order value were treated as fixed facts of life rather than things you could move. Tests happened occasionally, but nobody tracked what shipped, what it did, or whether the win was real, so the same debates repeated every month.

The approach.

I built a lab around the two levers that pay: conversion and basket size. A funnel view shows exactly where sessions drop off and where a test earns most, a blend-mix view tracks AOV against the bundle and free-shipping threshold program, and an experiment log records every test, its primary metric, its lift, and a verdict.

The outcome.

Conversion and AOV became things the team improves on purpose. The funnel shows where to aim the next test, the basket view shows whether bundling is working, and the experiment log ends the "did that actually help" debate. Winners roll to everyone, losers get reverted, and the lab keeps compounding.

+
INTERACTIVE SAMPLE

Explore the dashboard.

INTERACTIVE SAMPLE · Illustrative data, not client figures. Switch tabs for the funnel, blend mix (AOV), and the experiment log.

OPEN THE SAMPLE IN A NEW TAB
Guessing at what lifts conversion?

If your traffic is strong but the funnel leaks and baskets stay small, I will build you a testing system with a real scoreboard. One slot open for Q3 2026.

BOOK A 20-MIN CALL