Max Baasner
DE/EN
UPC · 2016Case study

−65% conversion in French-speaking Switzerland: A discount banner without context destroys trust

An undifferentiated partner-network banner reduced conversion in French-speaking Switzerland by 65%, showing that an offer without context can be interpreted as a risk rather than a benefit.

A/B TestingConversion OptimisationE-Commerce

The result in one sentence

−65% conversion in French-speaking Switzerland. A discount without context feels like a risk.

My role

Conversion Optimiser, UX Strategist, Test Manager, Hands-on Implementation

Worked with

Content Manager / Publisher · UX/UI Designer

01Situation

An exclusive benefit was promoted but not explained.

Partner-network promotion

An exclusive CHF 30 benefit needed more visibility.

Approach

A sticky banner linking to a generic landing page.

Weakness

Neither the banner nor the destination explained who the benefit applied to.

02Complication

How do you inform people without creating uncertainty?

Hypothesis

A prominent banner would increase awareness and therefore sales.

Risk

If message and context do not match, the result is mistrust rather than incentive.

Evidence from usability tests

Users without enough information tend to call rather than purchase.

03Approach

An A/B test evaluated separately by language region.

Version A

No banner, the control version.

Version B

Sticky banner with a general CHF 30 benefit linked to a generic page.

Analysis

German- and French-speaking regions were evaluated separately, which made the effect visible.

04Results

French version

−65% conversion at 97% significance

German version

−14% conversion at 55% significance

Cause

Unclear communication created open questions, leading to abandonment or phone calls.

Value of the test

A planned intervention was stopped before broad rollout.

Before · after
Before
Before
After
After
Methods
A/B TestingQuantitative Analysis
Tools
Adobe TargetAdobe AnalyticsAdobe Experience Manager

What I take away from it

Trust comes from clarity, not discounts

A benefit without explanation can look suspicious.

Analyse language regions separately

The German result alone could have hidden the real problem.

Test content, not just appearance

Since then I test message and context as systematically as layouts.

More projects
max.baasner@gmail.comEmailAll projectsArchive