Interview Prompt: PM for Meta - Design a product to recommend doctors
Should we build it?
If yes, how would you design?
1. Assumptions
US only
No constraints
2. Product landscape and motivation
Competitors
- ZocDoc - reviews, mostly scheduler
- Yelp - can review doctors, but based on personality and non-medical things; but don't know who the other reviewers are
- OneMedical - closed system
- WebMD
Company Mission: Build the future of human connection and the technology that makes it possible
Product Mission: Help patients find the right doctor based on their needs
- Trust based decision where people ask for recommendations through word-of-mouth and social proof
- Meta already has millions of health related groups (e.g., parents of SF, chronic pain support) where users ask for doctor advice
- With social graph and strong engagement in communities, Meta is well positioned to make doctor reco's more reliable, personal, and context aware
3. Customer segments
Patient
Life events - new parents- People share this online already so pretty natural
- Extremely active in Facebook groups already (e.g., local parent communities)
- Moving towns
- People share this online already and may not know anyone in the area
- Serious health issue - cancer
- Hard because people might be reluctant to share info online
- Chronic illness
- General adults
- Elderly
Doctor
- OBGYN
- Pediatrician
- Primary care
- Specialists
- cons
- Very hard to get ratings based on medical ability but could get other things (e.g., bed side manner, timeliness)
- Doctors don't want this; Lawsuits potentially
Insurance
Hospitals
Caretakers or other decision makers
4. Pain points
Lack of trustworthy, personalized doctor recommendations
- No trusted source: don't know if review is fake or reviewer shares your values
- Information overload: too many options online and hard to know who to trust
- Lack of contextual relevance: reviews don't account for parent specific needs (e.g., good with infants)
Non-medical dimensions very important
- Want to find a ped that meets your personality and bedside manner needs
- Want them to be available 24/7 via phone/text/messaging
- Want to be able to schedule last minute
Fragmented conversations: doctor reccs are buried in old facebook threads, hard to search or extract insights
Finding a doctor in a new area
Schedule an appointment easily
Confirm whose in-network insurance purposes
5. Solutions
GenAIDoctorFinder: users describe with NLQ what they are looking for and where they live….insurance, location,...filter by tech saviness, general responsiveness
- Could also put in medical problems and it could help give you some general WebMD type advice (but smarter) and help direct you to the right specialist
DoctorByFriends: Get recommendations from friends on who is good doctor vs bad.
Group Doctor Wiki: Each group can collaboratively create a trusted list of doctors (similar to Reddit "pinned recommendations" or Notion pages).
Facebook Health Hub (Long-term): A centralized tab for health help: find doctors, read advice, track appointments (integrated with Meta AI and local resources).
6. Evaluate
Pre-mortem: People can just use existing products and tools to accomplish this
7. MVP
P0
- Users privately list who their preferred doctors are but make it available to friends that ask
- One-Click Ask Feature: "Ask my network" feature that posts a templated question to your friends or a local Group ("Looking for a pediatrician in SF who's great with first-time parents").
- Could also just ask your network privately (without any messages being sent; just retrieve other users doctors)
- Could have automated prompts to respond and tag for future use.
- Filter by tags like "pediatrician," “OBGYN”, "accepts Aetna," "great with nervous kids."
- Could suggest group formations based on who provides info and who asks (e.g., parents with kids under 2, expecting parents)
P1
- Could add in a scheduling component with B2B office management for doctors so that the doctors have an incentive to use it and connect it with insurance
Success metrics
- MAU
- Hours spent on product
- Engagements (likes, posts, comments, etc)
- Hours / MAU