Question:  Design a Facebook feature to find roommates

Scoping: Is this feature going to be local to a market or is going to be available whereever FB is? – Global

Is it on the FB app? Instagram, Whatsapp, Marketplace?

– It is primarily on the FB app, but could use Instagram and Whatsapp as support application

Are there any other constraints that I need to know about?

– Both short-term as well as long-term roommates.

Can a third party advertise for someone? No only a person who has a ongoing lease can get new roomates

Goal – match the “correct” roommates. – product goal

– Connect people – FB mission

 

Customers (and their associated pain points):

– Poster (the one who posts the advertisement)

User journey : (1) Create an advertisement  with the approriate constraints and information -> (2) Publish the advertisement ->  (3) Deliver it to the right people -> (4) Interview roommates->  (5) Select roommate (s) ->  (6) Create finalize documents

Pain points for advertisers and potential solutions and their associated value – ( Impact, Technical, Cost, Time)

(1) – Mechanism to specify all the requirements for roommates – Location of accomodation, furnished/unfurnished, type of roommate – (gender, smoker/non-smoker, vegetarian, etc), rent information, utility information, sharing details, photos of accomodation, means of contacting advertiser (privately if needed). Feature interface that gives a standard list of requirements and allows the user to add specific requirements as well as upload photos (H, M,M, M)

(2) Publishing this on FB, Marketplace, FB groups, whatsapp lists etc. (H, L, L, L)

(3) Find the right subset of people to deliver this advertisement to.

– Using signals provided by potential roommates and find matches. Improve the friend finder algorithm to target roommates (H, M, M, M)

(4) Easy way to interview potential roommates

– Automatially setup video conferencing links once the poster. Can use messenger or Whatsapp – (M, L, M, L). Also helps avoid frauds

– Provide virtual reality walk throughs as well as mechanisms to place their furniture virtually to see if it fits (H, H, H, H) – definitely post-MVP

(5/6) Once selected enable digital signing of documents (L, M, M, M)  – Lower impact as other traditional methods are available

 

– Applicants

User journey:  (7) Set preferences for accomodation -> (8) Get a filtered list of accomodation -> (9) Connect with advertiser -> (10) Accept accomodation (11)  Sign documenation (12) Move to the accomodation.

(7)  Application that allows applicaints to give preferences (H, M, M, M). optionally publish this in marketplace also.

(8) Use signals provided by the posters to present a filtered list (H, M, M, M)

(9) Use whatsapp/messaging video to communicate with poster. (M,L, M,L)

(10) Digitally sign documentation – (L, M, M, M)

(11) Moving – Connect with movers advertised on Marketplace and get them to bid on your move – (H, H, M, H)

 

Of the above solutions I will choose the ones with High impact and lower cost , i.e. 1, 2, 3 and the corresponding ones for the applicant for MVP.

 

Metrics:

 

Number of people aware of this feature

Number of users using the service

Increase in engagement with groups, marketplace, messenger, whatsapp, instagram,

Number of successful matches.

Net promoter score

Design a Facebook feature to find roommates

 

Clarifying questions

• Before going to design a feature, can I ask you some clarifying questions to make sure we have the same thinking about this feature?

• I am not sure about the vision of Facebook. It is “Bring people together”, isn’t it?

(Assuming the answer is yes)

• Are there any target markets for this feature? For example, in America first.

(Assuming the answer is globe)

 

1. Goals & Constraints:

• From my point of view, this feature will have specific Goals like:

  • Bring people together (same as Facebook’s vision)

  • Collect more data

  • Become a supper app

• Please correct me if I lack anything.

(Assuming the answer is that they are adequate)

 

2. Users and User story:

   1. Users

We will potentially have various customer segments using this feature, but I want to narrow them to 3 kinds of users who have most significant impacts on our systems. I assume that my choice is made after doing market research and customer interview)

 

3. 2. 1. Students

• Students who have to study abroad or far from home and want to have roommates.

• Having a roommate help them save money, have a close friend, and support in case of emergency.

 

3. 2. 2. Lonely Travelers

• Travelers who travel alone and want to find a suitable accomodation for a period of time, more than half of a month.

• They don’t want to spend a lot of money on accommodation and they want to experience the local culture with their local roommates.

 

3. 2. 3. Employee

• Employees who have to work far from home for a short period of time that is not long enough to rent or buy a house.

Which one do you want me to focus on?

(Assuming the answer is Students)

 

3. 2. User story:

• According to me, there are 3 main user story we should focus on first:

3. 1. 1. First, as a student, I want to find the information about my roommate, so I can contact them and ask more about them before making a final decision.

      2. Second, as a student, I want to set a clear payment method with my roommate, so that we can negotiate in advance before making an agreement.

      3. Third, as a student, I want to rate and review my roommate after s/he leaves to make sure her/his information on Facebook is trustworthy for the next selection.

 

3. Use Case Prioritization:

• Because our resources (time and money) are limited, and I think we have to prioritize which feature we should create and test first.

• To do that I introduce some criterias to evaluate:

• 10 is most positive and 0 is most negative

 

 

• According to the overall point, I will pick the Find Information use case to implement first. Do you have any arguments?

(Assuming the answer is no)

 

4. Brainstorm Solutions:

   1. Forum

      1. Brief

• Users can create a topic to find their roommate

• Other users can come to the topic, write a comment and freely contact the post’s owners.

• Post’s owners can close their topic when they find suitable roommates.

 

5. 2. 2. Pros

• Users can take control of their contents.

• Easy to implement

5. 2. 3. Cons

• Users may be confused because of overwhelming data.

 

5. 2. Recommendation 

      1. Brief

• First, they should provide some personal information and roommate requirements to ensure our suggestion is suitable for them.

• The system will recommend a suitable roommate each time based on the user’s information. Users can see the detailed information about other roommates and contact them if they want. Otherwise, they can ignore and see another person.

• After meeting suitable roommates, they should turn off the “Finding” mode to avoid being contacted by the others.

5. 1. 2. Pros

• Avoid overwhelming data

• Avoid repeat

5. 1. 3. Cons

• Contents are fixed.

• Recommendation system may take a lot of time to implement and release

 

5. Conclusion

• At the first stage, I prefer the first solution, Forum. It may help students have more choices to find suitable roommates by reading individual posts because the system doesn’t have too much data to implement recommendation systems with reasonable precision that may make users go away and never come back.

• They can see more details about their potential roommates through their Facebook.

Design a Facebook feature to find roomates:

1. Clarify:

Would be designing this feature for Facebook only or in other platforms that facebook supports Eg: Whatsapp, Insta.? This will help me scope the problem for specific platforms. In this question we will restrict ourselves to only Facebook Web and App. We wont be designing a new app for this but will be a sub-feature within Facebook

When we speak of finsing roomates, are we considering only long term stays like shared accomodations/rentals or even short term stays like vacations? We would be keeping this open for now and see how to approach for different users types and prio at a later point.

2. Classifying the different users type:

Since this problem has both users from the demand and supply side. Given Facebook today is not in the business of real estate and doesnot have any assets in hand it would be safe to assume that Facebook wont be providing the real estate to stay. Rather it would be connecting users and providing the platform to both users who are looking to share their place (providers) and people looking for a place to stay (renters)

Besides we can also classify the users based on their need of a place to rent for short term (1 to 2 weeks) vacation rentals/travellers persona or longer term 1 to 2 years house rental persona.

3. What are the user pain points:

Renter:

• Difficult to easily connect with providers for a place to rent
• Access to pipeline of multiple options
• Limited information about the place and neighbourhood
• Higher prices because of commission fees for brokers
• Trust and safety of the place to rent is a concern

Provider:

• Difficult to easily connect with people looking to rent a place
• Limited information about the tenant or person booking the place
• On time and hassle free rent collection
• Capacity utilization in terms of room nights

4. Alternatives in market:

• Airbnb providing vaacation rentals, not yet in to long term rental business
• Some co-living companies are in market that provide long term rentals
• Facebooks strength:
  • Connecting 2B+ users and already bringing them closer socially
  • Strong user data already present as a starting point and building on top of it

5. Narrowing use-cases:

• Marketplace to connect providers and renters
• Build trust within the system for both sides
• Gathering more information about the listed property
• Building a pipeline of options for a region for users to choose from
• Help with payments
• Help with capacity utilization

6. Prioritization:

• #1, #2 and #3 use cases are critical for this service to come to existence and to build the MVP
• #4 will be critical shortly post launch and need to be addressed in mid term
• #5 and 6 are still good to have and can be tackled later

• Build a new news feed post type called rentals
  • Users can fill a preset form and generate a post for Rental ad
  • Users can control the privacy setting of sharing the post with friend or public
  • Option to broascast your post within a certain Geo as limited by the form fill
  • Facebook collects property licensing documents as a part of the initial info gathering to ensure the post is valid
  • Similar to commenting to a post a user can show interest in the property buy filling some intial details which will verify identity
  • Pros: This is organic, within the feed, geo specific
  • Cons: Cluttering the newsfeed, no option for users in a different geo looking to move to a new place to be notified,
• Build a new tab for Rentals similar to Facebook market-place
  • Almost all features that re built aaboive applies
  • Better filtering functionality can be built within the system wrt: price, proximity to public transport, markets etc.
  • More focused approach and intent driven users
• Mid to longer term:
  • Build a payment interface when users can add their cards and pay directly and receive rent receipts
  • Personalized search recommendations and push notifications for users showing intent for a particular area
  • Automated Rental posting for a aggreement due to end soon
  • Using VR to have a virtual tour of the place before booking

• Facebook can start with the low impact news feed option to judge interest in a pilot market
• If interest picks up they can start investing into the specific tab for it
• Start locally for a few markets and start expanding