Role
IA Program lead @ Doctolib Connect
Duration
5 months - Ongoing
Industry
HealthTech
Responsibility
Research & Discovery
Design Management
Project Management
Data Analysis
Workshop Facilitation
Team
Product Director
Head of Design
User Researcher
Data Analyst
2x Sales Managers
TL;DR
Doctolib acquired Siilo (now Doctolib Connect) to scale clinical collaboration. Two years post-acquisition, adoption was at 10% and the products remained siloed. I identified this as an architecture problem, not a feature problem, and initiated and led a cross-functional IA program to restructure how collaboration works across products, teams, and markets.
+15%
Care team messaging activation
+20%
Contact sync adoption
70%
WAU for new organizations
with 5-35 members
Shared
collaboration layer across the company
Disclaimer: Confidential information has been omitted or obfuscated. This case reflects my own perspective and not necessarily the views of Doctolib.
Context
Doctolib acquired Siilo to unlock collaboration at scale. But integration didn't happen!
Doctolib acquired Siilo to unlock collaboration at scale. Two years later, most Doctolib users never touched Connect, and most Connect users saw no improvements.
10%
adoption for Doctolib users of Connect app
150h
wasted per year per GP because of missing data source
20%
of patients repeat bloodwork due to missing data source
The product had been treated as a bolt-on messaging tool. What was missing was a structural collaboration layer that connected clinical tools, team context, and patient data across the platform.
Initial vision design vision for the Mobile app.
Objective
Build Structural alignment, and Position Connect within the organisation
I proposed an Information Architecture program to the Product Director, Head of Design, and Head of Engineering, and got the mandate to lead it.
The bet
If we restructured the underlying logic of clinical collaboration before shipping more features, we could stop teams from building in circles. We can achieve this by:
Defining the structural model
Aligning teams on shared beliefs
Building a system that absorbs growth without fragmenting
Deprioritizing our key segment
I deprioritized small practices, our most active segment. After the first discovery cycle, I saw that large organizations carry the structural complexity that would break our architecture if left unaddressed.
Small practices run linear workflows so they'd benefit downstream. I made the call to solve the harder architecture first and scale down, not up. This was contested internally.
Strategy starting point: Connect the 2 main pillars of the CC&C industry (Clinical Communication & Collaboration) along with Connect's USP to build the best Communication & Collaboration platform in Europe
My Messages Section Architecture explorations with David B. and Felix T.
System
A Cross-Domain layer that connects the system and leverages network Effects
I worked with the Design team to setup 3 design principles and 7 beliefs that now govern design decisions across product teams. Some of those included:
Thinking for Mobile, designing for Desktop
Every metric said desktop, but I chose mobile-first. Clinical collaboration happens between rooms and shifts, not at a desk. We think for mobile, design for desktop first.
Building a cross-domain collaboration architecture
Feature teams didn't want to integrate with Connect. I embedded with their designers, co-designed the integration points, and proved the layer works from inside their workflows.
Designing with confidence
From prior trust research I led, I knew confidence is the foundational lever. If professionals don't trust the system, adoption dies regardless of functionality. I made it a structural principle, not a polish pass.
From D.Connect as a separate domain
Doctolib Connect was treated as messaging tool, separate from other tools, teams, and patient contexts.
To Connect as a Horizontal Layer
Doctolib Connect became a collaboration system, where conversations enable shared context, teams and responsibilities.
Validated UI annotations
Impact
Care teams became more active. Product teams aligned without Friction.
Ongoing experiments to mobilize the system validated the direction:
+15% Care Team Messaging activation
+20% contact sync adoption
70% WAU for newly registered organizations that have 5–35 users
The larger impact
Product teams stopped building overlapping features. The IA initiative gave engineering, product, and GTM a shared structure for where Connect fits, reducing rework and setting the foundation for CareCo's expansion roadmap.
Learnings
Strategic design means containing ambiguity, not resolving it
Sustainable scalability can only be achieved by containing uncertainty so teams can move without constant redesign. Invest earlier in alignment rituals: smaller narrative tests, clearer tradeoffs, and explicit non-goals.
System redesigns fail without calibrated narrative
With systems redesign, narrative should be meticulously calibrated per audience. Concerns must be contained, to enable feedback instead of panic. Mistaking resistance for misunderstanding slows progress and costs trust. I learned to be deliberate about where to hold firm, where to adapt, and how to sequence decisions.
