AB Testing in Livesport

Case Study

AB Testing in Livesport

Optimising conversion through data-driven experimentation

CRO Specialist2024

Role

CRO Specialist

Timeline

1 Year 3 Months

Platform

Desktop / Mobile

Tools

FigmaFigma
OptimizelyOptimizely
GA4GA4

Overview

The Challenge

As a UX-focused CRO specialist at Livesport, I was embedded in a high-traffic product with millions of daily users across Europe. My role sat at the intersection of data analysis, UX thinking, and rapid experimentation — working alongside product managers, engineers, and analysts to test hypotheses and validate UI changes through controlled A/B experiments.

Case study detail

Research & Discovery

Key Findings

Finding 01

Small visual tweaks — rewording a CTA or repositioning a badge — often had a disproportionately large effect on CTR

Finding 02

97% of web traffic came from mobile, meaning desktop-first assumptions consistently undermined test validity

Finding 03

Positioning CTAs near live-event or betting elements increased engagement by up to 34% in controlled tests

VIEW ALL

Design Process

How We Got There

01
Define
02
Research
03
Analyze
04
Design
05
Test

Each experiment began with a clear hypothesis grounded in observed user behaviour. I sketched multiple variants — from microcopy changes to full component rearrangements — then narrowed to the two or three most promising candidates for development. Design decisions were always tied back to a measurable metric.

Variants were built in Figma and handed off to engineers as interactive prototypes. I wrote the front-end implementation specs and collaborated closely during QA to ensure pixel-level accuracy. Post-launch, I monitored statistical significance and documented learnings to feed the next cycle.

VIEW ALL

Final Design

The Solution

Final design
#1

+18% overall conversion uplift across the tested funnel

#2

−31% drop-off rate at the seat-selection and match-detail steps

#3

−22 s average checkout completion time after layout optimisations

Takeaways

Reflection

Microcopy and progressive disclosure consistently outperform full visual redesigns when iterating on a mature, high-traffic product. The key discipline is measurement rigour — every design decision earns its place through data, not intuition alone. Small, validated bets compound rapidly at scale.

Final design detail
NextWireframe Design System
+
+
+
+
0000 · 0000
Daniel Gratza

Let’s make something great.

Prague-based UX designer

Get in touch