Question

Design a grocery app for Facebook?

Answer 1

Identify the North Star and alignment to Facebook's mission. Ask questions to understand what the interviewer is expecting to you to build.

• Grocery shopping is something people do routinely. They share advice on products, mainly food (recipes) with friends and family. What do you imagine this app helping with? Shopping? Recommending grocery items? Sharing recommendations with friends? etc?
• I think it makes sense to leverage the audience and features integrating into Facebook vs. a standalone app. Is this the direction your thinking? Why?
• What's the primary platform for this use case, desktop, mobile, watch, VR, API? Why
• I think the main value for FB here is to leverage relationships to find better products, products are broad, narrowing on food seems core to what people have a need to share. Also leveraging existing features in Facebook like photos, videos, comments can really help. What are your thoughts on narrowing the focus to food in the MVP? Are there other categories that you see as relevant? What do you see as the main value
• The best way to share grocery items with friends & family.
• I think the main metric could be how many people repeat share and frequency of shares, what other measures of success do you see?
• Will users be able to purchase groceries? 
• Are users sharing to specific people or broadcasting?
• Where is the primary location? Village in rural India or NYC?

Identify the user. Personas

• People having a need to share grocery (food) items they care about. 
  • Should we narrow to busy professionals? Male/Female? Parents? Mobile Facebook users? Age? What are they sharing? Why?
• People receiving recommendation
• Grocery advertisers? 
• Parents who both work, they use the Facebook mobile app daily, 24-35 years old, and have at least 100 friends on Facebook. As the success metric goals are reached, the feature can evolve to more personas.

User Stories

1. As a user, I can take a photo of a grocery item and share it with their network.
2. As a user, I can see what items are recommended by my network.
3. As a user FB applies smarts to personalize and prioritize recommendations then displays to me.
4. As an advertiser, I can set a grocery inventory, price, add, and pass to FB that a user might recommend to their network.
5. Users will be able to see nearby grocery stores with the item
6. Users will be able to purchase the item online

Prioritize

1, 3, 4, 2, 5, 6

• Rationale: 1 is a predecessor assuming a friends recommendation is valuable. 3 would have a big impact on scaling, saving time, and making the feature value better than the alternatives currently on the market.

Draw a Wire Frame detail for 1.

Measures

• # of shares
• The frequency of shares amongst unique users/repeat rate
• CVR of viewers
• Revenue from advertisers

Recap

We want to build the best grocery recommendation feature on the market that will be integrated into the FB app. Targeting busy professional parents between the ages of 25-34 in major cities. The MVP will have a simple interface for a user to add a photo and share with their network, it will look like x. Viewers of the recommendation will be able to see a nearby spot to buy and can even order with their device and pay. We will measure the number of shares, the frequency, the repeat usage, CVR of viewers recommended grocery item, and revenue from advertisers.

Answer 2

Clarifying questions:

1. Is this app going to help buy grocery or provide recommendations for items from your network?
2. Is this for any particular region?
3. Any specific grocery category like baby products, organic produce etc.?

Considering this is a new feature I would select an urban city like SFO where there are more working professional and launch MVP. Once I gather feedback and findings from launch, I will decide to expand this in other cities. 

For FB to launch a grocery app, we need to understand its competitive strength on this area. Often one relies on friend’s feedback to try any new product or baby products/food items. FB social graph can really provide a great value to one when exploring new items, deals, organic produce or baby products. Also, FB can really help connect one to produce from local farms which often requires one to rely on feedback

Product Goal: Shopping for grocery is a time consuming task. Parking queue, crowds and standing in queues for checkout is very time consuming. Online grocery app will thus provide the convenience of shopping from home and remove burden of standing in queue

Business Goal: FB as one stop place including grocery and increase engagement by time spent on FB

Users:

Users of this product would be tech savvy who often like to buy online. Of these below users come to mind

1. Students in universities: Stay with roommates. Busy with courses and exams and do not have time to go to stores
2. Young professionals /Couples: Busy professionals /couples who do not have time to go to grocery store. They would not like to spend weekends on these chores
3. Families with kids: Busy with kids but need to make more frequent trips to store due to kids

Out of these would like to focus on young professionals and families as they have more disposable income than students to buy online and pay for potential delivery cost. Also students mostly end up eating fast food in university cafe so may not buy grocery often

Pain Points

Some of the pain points of in store shopping are to make the grocery list before hand every time. You also need to share the list with your spouse in case the spouse ends up going to the store. Even after that you forget some of the items.

If you are planning to make any new recipe then you need to make a list of all ingredients before you go to the store. Also often not everything is available in one store for e.g. you need organic produce that is available in one store and other items in different store and your grocery chores ends up visiting multiple stores which is very time consuming. Lets address some of these pain points

Use cases:

1. User can choose items from variety of stores in the local area so there is no need to visit a physical store. Easy and Fun to tap icons for familiar food items to add them to the list.
2. Allow one to search for and item and price compare an item from multiple stores
3. Suggest local farm produce based on one’s network
4. Allow one to easily make a list & share the list with family members
5. Allow one to add notes and comments to shopping lists.
6. Remember previous week's list to help with new list.
7. Allow one to add all the ingredients from a recipe to the shopping list with one click.
8. Calculates the total spending amount of the list and offers suggestions for alternatives stores within the inventory to reduce cost
9. Matches items in weekly ads and coupons with items on shopping list
10. Notify one when a product's price in the list dips below a set price.
11. Allow one to see the reviews on food items from within your social graph 
12. Allow one to see weekly the most bought item and the store within your social graph

Prioritize use cases

For MVP, I would focus on 1,2,3,4,6,8,11,12. since these are the basic needs while one is grocery shopping. FB can really help to connect people with local stores and organic farms and help them get feedback from their social network on the most commonly bought item. Often one is hesitant on sampling produce from local farms unless someone recommends it to you. Also, tracking deals from multiple stores is very time consuming and FB can really personalize the items based on what I frequently buy based on the deals and recommendations.

 Solutions

1. FB can have a grocery tab where one can see inventory from local stores and fresh farms. User should be able to easily tap on items and add in the list.
2. User can create a list by manually adding the items.
3. Manual entry may be time consuming so one can also use voice entry to save create and save it.
4. User can price compare an item from various stores. This may be time consuming. Alternatively FB can automatically show you the top 3 stores based on price, reviews from social graph
5. Showing reviews on products from your social graph. User can see comments, reactions on the items from social graph
6. Using AI, FB can see the most frequently purchased items within the list, provide you recommendations on where to buy that from based on feedback from your network and provide you the best deals based on the weekly ads

 We will prioritize the solutions based on impact, cost and complexity

1.  High, Medium, Medium.. FB needs to partner will local stores and farms to have their inventory added in their website
2. High, low, low – Manual entry is basic need for grocery shopping
3. High, High, High. Busy professionals can create this on the go using voice than manual entry
4. High, High, High – FB to automatically show you the top stores based on price and review. This may be very complex to begin with so MVP will launch with users manually adding the stores for price comparison
5. High, High, High – This provides high value to user as often you rely on feedback to try new items or stores
6. High, High, High – This provides high value to user but for MVP integrating deals that matches your list will be very complex

Thus for MVP, FB can design a mobile app focused for a city like SFO that will have inventory from multiple local stores and farms. Once can easily create a shopping list manually or via voice, price compare and see the reviews and reactions from social graph.

Answer 3

Design a Grocery app for Facebook:

Facebook has its mission - Give the people the power to build community and bring the world close together. There are multiple facets of our lives around which communities can be built. Grocery shopping is a routine weekly or multiple times a week task for most adults and people can share useful information with each other and connect over grocery shopping.

Will it be an app on FB or mobile app or a website? (assuming it would be FB app)

Will the goal of this app would be to drive more user engagement, business revenue, improve a sense of community among users?

I am taking the goal to be improving a sense of community if thats ok. For this, Grocery app for improving a sense of community can be either an app where people recommend stores where rare to find things are available or stores which are really good for somethings. Also share special grocery items for specific groups like kids, religious practices, old aged, etc.

I would now define a set of users and determine which would be benefitting the most, then identify their needs and based on prioritized needs, define which solutions could be designed to meet those needs.

There can be many potential user sets:

1. Pregnant ladies or Moms of kids upto 3 years 

2. Old aged people above 60 years

3. Health conscious people

4. Busy professionals who want quick and health recepies and their grocery items

5. Specific communities which have religious ceremonies or festivals

I would want to pick up a user set which would be benefited maximum by continuously contributing as well as getting information from the community and would be substantial in numbers.

I believe that health conscious people would be a group i would want to focus on and there are increasingly high number of such people. Also these users have evolving needs of beginners, advanced, specific to different sports or activities as well as different diets. Thus a lot of value can be created through this community.

These users would have the following needs:

1. Share their activity or sport and interests details as initial data( put on weight, reduce weight or specific dietary requirements like reduce carbs, specific disease)

2. Can join subgroups based on their requirements 

3. Can ask questions 

4. Can have a session with an expert or answer a survey to find what is recommended for them

5. Can recommend foods and stores to find them based on their requirements 

6. Can attend live sessions by experts

7. Would want them to be delivered

I would prioritize- 1, 3, 4, 5 to start with in order to understand if there is a meaningful community that can happen around this and then deliver related groceries

Solutions: 

This would be an app like marketplace, offers, recommendations which user can have access to through Facebook and as per FB algorithm would show up in the organic feed of the user at times.

1. The user enters basic details like gender, goal, activity, any disease or deficiency, allergic reactions so that suggestions can be personalized, user will be given suggestions per day on the kind of recommended recipes and groceries based on initial data from a database which is initially added by FB based on inputs from experts. User can then later ask the questions from the community about recommended foods and stores where they are available. Also user can ask questions from the expert or book a session with them. 

2. The user can fill in basic details like gender, goal, activity and then can follow some recommended expert profiles. User can go through posts from those profiles and engage by asking questions or answering some. A user can share their experience/recipe per their specific needs. Recommended recipes can show up at the top or through push notification each day if enabled to the user which will have list of groceries and stores to purchase them from. Experts can be engaged by Facebook as the users can also book sessions with them in return.

3.  The user can share in basic details like gender, goal, activity and based on their dietary requirements, user can generate a weekly chart so that it can follow for a comprehensive diet over the month. Users can ask for questions and recommended grocery stores. Over time this data can be processed through NLP and user can be recommended grocery stores for their weekly charts. This will leave more room for the user to ask questions about deeper things like specific to a disease or allergy, sharing experience of buying at a particular store, etc.

I would now evaluate these solutions based on user value, risks, and engineering effort.

1. user value- medium, risks is low as the data is recommended through experts input, engg effort- med as decision trees for the recommendations or recommendation engine is to be developed which will learn and become better for the user.

2. user value: medium, risk is high that in order to engage the experts initially can be challenging so as to generate some content which a new user can engage with, engg effort: medium- recommendation engine

3. user value: high as now user can goto FB everyday and keep better track about their health through diet charts. risk: medium as not all users will be engaged to the diet charts as well as optimizing NLP so as to be solid enough to start answering user questions based on data gathered, engg effort: High

So i would prioritize 1st where some experts can be engaged or data can be pulled up and organized to start with before the system becomes attractive to experts on its own. Doing MVP with this even though 3 has high user value because its cost of implementation is high and risk is also high.

After we see the initial response based on user feedback- weekly charts, NLP auto answering can be further developed to enrich community experience.

Thus, in order to increase sense of community, i would create a solution that would allow users to ask questions about recipes and grocery stores as well as get recommendations from experts input on daily basis.

Comments

Hi srocks,
Thanks for posting your answer. I have some feedback on this:
- User groups / personas should be exclusive. e.g. students, young professionals, families with kids. Some people might belong to more than one group but at a high level, they should be exclusive.     
- Each use case should reflect a need. The listed use cases need further explanation. For example, why are "joining subgroups" or "asking questions" important use cases?
- Keep each solution description brief and ensure that they each reflect a unique feature. Check out the examples submitted by Ali or prisingh above as reference.    
- I would list at least 5 solutions for product design questions to ensure that you can pick a minimum of 3 as the key features of your MVP
Hope you find my feedback useful.

---

I really liked your approach. Your solution is to build an app which would provide personalized dierty plan and also recommend where to buy the grocery from. I do agree with Bijan that the above solutions are not distinctive enough, they are kind of following the same approach. I think its because the needs are very generic. You picked a persona "people who are health conscious". Pick up their pain point and identify solution based on that. Good luck!