01
The referee
Assesses how urgent a symptom is before any appointment.
Sector expertise — Health & Wellbeing
Your patients have a
doctor number 2.
Do you know what it's
telling them?
For 10 years, we have been doing field research with patients, carers and healthcare professionals to keep understanding their needs better and serving them well. More recently, we have been studying the impact of AI on care pathways.
Signal · unknowns ethnographic study · 2026
Near-automatic reflex
The second opinion
Even before seeing a doctor, AI has become the first reflex: describing symptoms, getting a prescription checked, asking AI to decide how urgent it is.
Recurring behaviour
The three-way consultation
Patients arrive — or leave — with ChatGPT's answers. The doctor-patient face-to-face has become a triangle. The appointment is no longer a duo.
Already observed, high stakes
The silent switch
Some patients adjust or stop their treatment on AI's recommendation, without ever mentioning it to their healthcare professional.
Source — unknowns qualitative study: 12 in-depth ethnographic interviews, 2026. Qualitative observations, not statistically generalisable.
Our experience
10 years in the field
in healthcare.
Studies of patients' lived experience and doctors' practices, mapping care pathways, scientific publications, designing new services and even clinical tools, and now bringing AI into pharma marketing: we work on issues that directly affect patients' quality of life.
- Roche
- Amicus Therapeutics
- Novo Nordisk
- Juvise Pharmaceuticals
- Société Française de Rhumatologie
- INCa
- Servier
- UCB
- SOBI
30+
Qualitative and quantitative studies in healthcare
500+
Patients interviewed in our research programmes
100+
Healthcare professionals met through immersion
50+
Scientific publications co-authored with medical teams
AI Observatory — Health module
A long-term research programme.
Dedicated “Health & chronic illness” module — data updated twice a year, available through sector co-funding.
In the field
What patients
really say.
The doctor gives me a medicine… ChatGPT explained exactly what the medicine did… So, well, I didn't take it.
It's really become my doctor number 2. I don't even ask my husband questions any more, I go straight to the AI.
I think I use Google half as much. AI answers my situation, not a generic question.
5 roles
AI is not one thing.
It plays five distinct roles.
Each role comes with a behaviour, a risk, a business opportunity and a lever for your teams.
02
The translator
Decodes medical reports, test results and jargon.
03
The teacher
Prepares the patient, rebalances the information gap.
04
The counter-expert
Challenges the prescription, backs or overrules the doctor.
⚠ Critical risk for pharma
05
The coach
Provides day-to-day support in managing a chronic illness.
Key insights
Three tensions reshaping
your business.
What we see in the field: a deep need, an established behaviour, an unaddressed business risk.
Deep need
“To be seen as a person, not a file.”
AI is seen as empathetic and available. It listens without judgement.
A need for a more equal footing: becoming an expert in one's own illness.
Behaviour
“The patient-doctor relationship has become three-way.”
Patient, healthcare professional, AI. The triangle redefines medical authority.
Near-systematic behaviour: showing AI's answers to one's doctor.
Business risk
“AI-driven non-adherence is already here.”
Patients check their prescription against AI and quietly adjust.
Already observed: changing one's treatment on AI's advice, without telling a carer.
Offer
Strengthening medical detailing
through patient insight.
With fiercer competition and AI erasing brand names in favour of INNs, differentiation comes from the most detailed understanding of patients' lived experience and doctors' practices.
01
Immersive qualitative study
In-depth ethnographic interviews. Real drivers of prescription, lived experience of the illness, latent needs.
02
Identifying patient profiles
5 to 10 salient profiles per condition. Aspirations, key tensions, arguments in favour of switching.
03
Marketing & field activation
Detailing scripts enriched by profile, team training, AI agent that simulates consultations.
04
Scientific publication
SRQR guidelines, scientific committee. Abstracts, posters and oral presentations for medical congresses.
Deliverables
- Structured persona sheets
- Detailing scripts by patient profile
- Key messages for HCPs & patients
- Multichannel content strategy
- Congress abstracts & posters
- AI consultation-simulator agent
- Field team training
JUVISE PHARMACEUTICALS
Bringing AI into medical detailing
and pharma communications.
Teams spend precious time on repetitive tasks — LinkedIn posts, detailing scripts, agency briefs, handling objections.
- 01
Mapping real-world uses
Identifying what can genuinely be handed over to AI in a pharma context.
- 02
Generating and selecting prompts
5 to 10 candidate prompts per use case, tested on real past deliverables.
- 03
5 use cases covered
LinkedIn posts, agency briefs, detailing scripts, review checklists, objection-handling guide.
- 04
Phase 2 under way
Personalisation by HCP profile, scientific paper summaries, search across CSRs.
Outcome
A library of ready-to-use prompts across 5 immediate use cases. Agile approach in 2-week sprints — deliverables usable straight away.
Opportunities
Four engagements we could
take on together.
- 01StrategyRemapping the patient relationship in your therapeutic area.
- 02FieldGiving your field teams the right stance with doctors again.
- 03ProductDesigning the digital health service your patients will actually use.
- 04Internal AIRolling out AI across your marketing and medical teams without it staying a POC.
FAQ
What we're asked
most often.
By starting from how teams actually work, not from the technology. unknowns designs tools that support concrete uses: detailing scripts enriched by patient profile, prompt libraries for marketing, AI agents that simulate consultations… All grounded in a detailed understanding of patients' lived experience and doctors' practices.
Because AI thinks in categories and molecules (INNs), not brand names. When a patient or doctor asks an AI assistant, the brand often fades behind the generic. For a pharmaceutical company, differentiation now plays out elsewhere: in understanding patients' lived experience and in the quality of the relationship with doctors, ground unknowns has been working for ten years.
Patient insight is a detailed understanding of how people live with their illness, what really drives treatment adherence and what they expect from carers. unknowns builds it through ethnographic research. Now that AI is turning medical information into a commodity, this is the asset that strengthens medical detailing and differentiates a brand for the long term.
unknowns is neither a healthcare communications agency nor a tech provider. We combine social sciences, design and generative AI to start from the real behaviour of patients and carers (not from product messages). This field understanding, built over ten years in healthcare, is what makes our tools genuinely adopted.