The Problem

For our patients, reaching out to their doctor typically wasn't a casual chat, they often came at a moment of stress or doubt. Their skin may be reacting badly to treatment, they worried treatment wasn’t working, or they had big concerns like pregnancy, other medications, or how their skin would tolerate new products. Many looked through FAQs first, but when they finally messaged, it was because they needed personal reassurance from a doctor.

The original message feature design didn’t meet them in that moment. It mirrored a rigid consultation model:

• Patients had to open a new request, pick a theme, add photos, and then wait 24–48h for a reply.
• Once submitted, they couldn’t follow up or add more context.
  
• Doctors would reply and then “close” or “keep open” cases depending on perceived need, and patients didn't have any influence here.

For doctors, this queue-based structure worked well, it kept triage efficient and manageable. But for patients, it felt transactional and cold. The system put walls up at exactly the moment they needed connection. Over time, that mismatch eroded trust (which we see through reviews patients leave on individual message interactions), reinforced negative experiences many already carried from traditional dermatology, and drove cancellations before treatment had time to work.

Insights from Patient Data

The Challenge

Patients wanted continuity and reassurance; doctors needed structure and prioritisation. A thread-like chat promised more trust and engagement, but introduced risks:

• Reopened cases could flood queues and bury urgent issues
• Multiple doctors might respond in the same thread, breaking continuity of care
  
• Long, unstructured threads could slow triage and make safety-critical issues harder to spot

Chat Thread vs. Consultation

One of the biggest pain points was patients couldn’t reply or add further context once they hit send or a doctor “closed” a consultation after s reply. Many had follow-ups but were blocked, leading them to hack other channels.

Designing for Safety and Relevance

The new messaging flow wasn’t just about making it feel smoother for patients, it also had to work within the realities of medical care. That meant every change needed to balance patient clarity and empathy with doctor triage and safety requirements. Two key ares of communication we needed to address were themes (for categorisation) and side effects (for patient safety).

Learnings and Next Steps

We launched the redesigned flow with a small test group before rolling it out more widely. Adoption and engagement increased significantly, especially in the first month. Patients who messaged their doctor early were more likely to stay, validating our retention hypothesis.

But the flip side was an unprecedented spike in message volume: from 7.8k → 11k monthly messages. Doctors flagged new problems:

• Long “chatty” threads made it harder to extract critical info.
• Some patients treated it like WhatsApp, sending incomplete or off-topic messages.
  
• Prioritisation in Kabinett became less reliable.

Through workshops with medical, we introduced guardrails that balanced this new trust with efficiency. Quick replies that didn’t open new cases, limits on re-opens, and UI nudges like larger input fields to encourage more complete messages.