Handled during the period analyzed.
Case study · EIMEC · Medical training
Filter before you respond. 20,684 messages.
How we prepared every conversation to help EIMEC gather model patients without overwhelming the team that had to assess each case.
Active campaignMeta → WhatsApp
EI
EIMECModel patient assistant
Hi. Before I put you through to the team, I need to ask you three questions.
Are you currently in Barcelona?
Yes, I’m in Barcelona.
Great. Which treatment are you interested in?
Conversation readyTransferred to the EIMEC team
Simplified example
Barcelona · Late 2024 / Early 2025Scroll to follow the journey ↓
Conversations started on WhatsApp.
Reached the team with context already gathered.
Messages exchanged in each conversation.
The challenge
450 patients. A very short window.
EIMEC needed to gather about 450 model patients for a training course. Recruiting them too far in advance weakened commitment: many people still didn’t know whether they could attend.
Training goal≈450
model patients
Approximate period · 2 monthsThe peak came at the end
15–20 days
PreparationActive campaignTraining
When recruitment peaked, every ad led to WhatsApp. The bottleneck was no longer attracting people but being able to attend to them with good judgment.
The journey
One way in. Four steps.
The system didn’t try to handle the medical side. It organized each arrival so that every conversation started with the minimum information needed.
01 / CampaignsMeta and email→
02 / Entry pointWhatsApp→
03 / QualificationThree questions→
04 / DecisionEIMEC team
Are you in Barcelona?
Check that the person’s location was compatible with the training.
Are you familiar with aesthetic medicine?
Understand their starting point before continuing the conversation.
Which treatment are you interested in?
Give the team the context it needed to assess each case.
Judgment stayed human
We didn’t automate the clinical decision.
Deciding whether someone could take part as a model patient required a person. The automation handled everything before that: asking, sorting and transferring.
The team repeated the same questions before it could assess the case.
The team received the location, the context and the treatment of interest.
The rebuilt dashboard
The volume, explained clearly.
We recreated the operational metrics in this case study to show what happened without publishing conversations or patients’ personal data.
Executive summaryModel patient recruitment
Period analyzedAn average of 27.47 per session.
Conversations recorded in the dashboard.
69.72%transferred
Human follow-up
525 conversations
reached the team with the earlier answers.
Training outcome≈450 model patients
EIMEC managed to gather roughly the number it needed. A significant share went through this flow, although we don’t attribute the result solely to the automation.
The real result
The team responded better because it didn’t have to start from scratch.
Useful automation doesn’t replace judgment. It removes repetitive work so people can apply their judgment where it really matters.
See more projectsAggregated data · No names, phone numbers or conversations · Campaign of about two months · Intensive phase of 15–20 days