← All work

Removing friction from a product quiz

A UX audit that uncovered 15 small problems affecting completion.

UX AuditCROQuiz
View Figma exploration ↗
Problem

The quiz told users there were five steps but asked eight questions, and only 24.63% of people finished it.

What I did

I audited the flow, found 15 issues across four areas, and fixed expectations, action hierarchy, redundant questions, and mobile readability.

Result

Completion rose from 24.63% to 35.80% (+45%), and subscription revenue per visitor increased 3.93%.

Role
UX Designer
Audit and redesign of the quiz flow
Client
Happy Mammoth
Via DTC Pages, CRO agency
Platform
Web
Multi-step quiz, mobile-first

Context

The quiz asked users to answer eight questions, but the interface told them there were only five.

Several questions also repeated information or behaved differently from what the interface suggested. Completion was only 24.63%.

The opportunity was not to redesign the quiz visually. It was to remove the small points of friction that made the experience harder to complete.

Friction was everywhere, but nowhere was catastrophic.

The audit uncovered 15 issues across four areas.

Progress & expectations
The step count did not match the actual number of questions.
Logic & content
Some questions repeated information or contained inconsistent answer rules.
Interaction
Some controls suggested the wrong behavior.
Hierarchy
Important actions competed visually with secondary information.

What I changed

Here are four of the fifteen fixes.

01
Fix expectations
Correct the progress information and eventually remove the misleading step count.
02
Clarify the primary action
Make "Next" dominant and demote "Back".
03
Remove unnecessary questions
Use behavioral data to identify and remove redundant steps.
04
Improve mobile readability
Increase question size and adjust spacing for smaller screens.

Outcome

+45% quiz completion

Completion increased from 24.63% to 35.80%.

Overall drop-off decreased by 15 percentage points, and subscription revenue per visitor increased by 3.93%.

What I learned

Most UX problems are not one big broken interaction.

They accumulate through small mismatches between what the interface says and what it actually does.

In this case, fixing those small inconsistencies was enough to move completion by 45%.

Hiring for growth or e-commerce design?

I’m open to product design roles in DTC, e-commerce, and SaaS. Email is the fastest way to reach me.

Next: Designing an AI-assisted workflow from ad to landing page → ← All work