Tina El Hawa.
Back

Optimizing subscription conversion through address auto-complete (2025)

The new address step, with one search field instead of six separate inputs.
Company
Company: Free
Type
Type: Business optimizations
Product
Product: Self-service kiosk
Focus
Focus: Data & UX/UI

The address entry requires users to manually fill six separate input fields on a large vertical touch screen—a high-friction task.

Context

Our self-service kiosks had six separate input fields to ask the user their address info. This required us to add an extra step to the process, a “contact” step with name, last name, email etc. inputs and another step “my address” dedicated for these six fields. We knew it was obviously a friction but didn’t know how to prove it with numbers since we didn’t have any analytics tools for kiosks flows.

The goal: Leverage web-based insights to justify and implement a streamlined checkout experience on self-service kiosks.

The challenge: Free’s kiosks lacked behavioral analytics tools that were, however, available on our web platform. This created a "data blind spot" for in-store user behavior.

My approach

I acted as a bridge between platforms, using a "Data-Transfer" methodology to validate UX improvements where direct tracking was unavailable. For that, I worked hand in hand with the Product Designer in charge of the web platform and my PO.

The legacy address step, with six separate fields under Mon adresse.
Six input fields legacy which had its own step “Mes coordonnées personnelles”
The replacement step, with personal details and a single Adresse postale search field.

Web A/B testing: We ran an A/B test on the web platform comparing the 6-field legacy form against a single-field "Smart Address" auto-complete solution. The single-field version led to a +36% increase in users reaching the next step.

Technical implementation on kiosks: From what we learned thanks to web data, I worked with the kiosks engineering team to adapt this single-field logic, merging "Personal Info" and "Address" into one seamless step.

The result: I conducted on-site shadowing and interviews with shop advisors to observe the new flow in real-time. The estimated impact, based on field observations, was a 3-minute reduction in total subscription time per user.

Outcomes

Process efficiency: By removing the legacy 6-field system, we significantly reduced the "Time-to-Task" for in-store subscriptions.

Strategic cross-platform win: Demonstrated how to use cross-platform data to drive high-impact changes in environments without native analytics.

Bonus

Our kiosks were on TV ➡️ take a look here ⬅️

A Free kiosk plan screen filmed for television, with the M6 Info logo in the corner.

How I work for kiosk screens 🕹️

Two Free self-service kiosks side by side, the production replica on the left and the test kiosk on the right.
A vertical monitor turned like a kiosk, showing the Free offer screen next to the design file.

The kiosk on the left is the replica of our live production and the one on the right is our test environment.

I can also access the pre-prod with a localhost link on the web. On top of that, I turn around my PC so it has the same orientation of the self-service kiosk, really helpful when designing on Figma and check the look and feel.

More projects to see