As a PM for Google Maps, how would you improve the restaurant search?

Question

As a PM for Google Maps, how would you improve the restaurant search?

Asked at Google over a year ago

6.8k views

Product Improvement category

& 1 other company

Asked at

& 1 other company

6.8k views

How to answer Product Improvement questions

Interview Guide

Top Google interview questions

What is your favorite product? Why?89 answers | 263k views
How would you design a bicycle renting app for tourists?62 answers | 82.5k views
Build a product to buy and sell antiques.54 answers | 66.8k views
See Google PM Interview Questions

Top Product Improvement interview questions

How would you improve Google Maps?53 answers | 228k views
How would you improve YouTube?29 answers | 81.3k views
How can you improve Facebook Stories?22 answers | 45.5k views
See Product Improvement PM Interview Questions

Invite members
Invite by email

Add another

Cloud Strife · Answer 1 · 2021-01-09T00:37:10+0000

**Please note that I try to answer these questions as if I was in a real life interview. Thus they are usually short and condense (as oppose to a thorough exhaustive response). Interviewer responses will be in italic. I will also put my internal thinking process in [brackets].

Clarify

Great question! You’re asking me how I would improve restaurant search, correct? As in, I’m searching for a restaurant I want to eat at?

Yes that’s correct.

So before I jump in I want to think about this question and maybe ask you some clarification questions that might come up. Does that sound good?

Sure, let me know if you have any questions.

[Always ask clarifying question, no matter how simple the question might be. Remember the interviewer is more interested in your thought process than the actual answer you give. As a note, think about the different sides of the market/business. For example, the consumer side vs the business side]

When I think of restaurants search I can think of a couple thing

I’m in town, either somewhere new or where I live, and I’m either in my car or walking around and I’m searching for a place to eat (literally physically looking for a place to eat)
I’m on my mobile device or computer and I’m looking for a place to eat (mobile/web search).
I could be by myself, with a couple of friends, or maybe it’s a group of us (for a party or celebration)
Maybe I’m at home or work and want to order some delivery or takeout.

Also, as I was listing the above uses cases, It popped into my mind that maybe businesses or organization might want to find a place to eat:

I could be a business or organization that is looking for place to eat with my team of people (large organization, 20+)
- I could also be a business that is looking to a host an event at restaurant.
A business/organization that is looking to have lunch catered to my business
A business/organization that is looking to have lunch delivered (large order) to my business

So far, I’ve brought up restaurants search from the consumer/buyer point of view. What bout the restaurant side?

For example, if I’m a restaurants owner I might…

Want to figure out how good my search/reach is (SEO) with my restaurant and how to improve it.
- This could apply to my personal website listing or my listing on other websites (Yelp, trip advisor, etc…)
I want a better way to list and manage t my food, address, hours of operation, special, restrictions due to COVID. So when a potential consumer finds me all my information is up to date

My questions are:

Are we looking at solving the search problem for individuals or large business/organizations?
Do we want to include catering business or just dine in?
How about delivery?
Are we interested in the consumer side or the restaurant side?

Hmm. These are all very good questions! Given the current market, which problem do you think might be a good one to tackle?

Well, due to COVID, a lot of individual are working from home. Which means business are probably less likely to have employees working inside their offices.

So targeting business/organizations wouldn’t be beneficial. A LOT of people are ordering food delivery for a number of reasons: They are working from home, they are covid conscious and want to minimize going outside and being exposed, they want to save time by not leaving the house. From the restaurant stand-point, I don’t have a lot of personal experience so I’m not sure what exactly could be improved on.

So I think targeting search on the consumer side would probably be ideal.

I think that makes sense. So let’s focus on the consumer side.

So to clarify, we are going to focus on restaurant search from the consumer side of view.

That’s correct.

Great. So now that I have a better understanding on the problem we are going to solve, I want to give you a brief outline on how I’m going to tackle the problem.

First I’m going to list some potential users/personals. Next I’m going through the needs of one of the personas to narrow down the problem I’m solving I’ll then select one of the customer problems/needs to solve for. I’ll brainstorm some solutions and prioritize them. I’ll then figure out what’s the best solution to move forward with based on some attributes. Finally I’ll summarize my answer.

[Note that I am using the CIRCLES method to solve this design problem. So while I’m outlining my approach I’m not explicitly telling the interviewer each step within CIRCLES. You do not want to sound like a robot when you go through this process]

Does that work?

Sure, seems like a reasonable approach.

Great. Then next I’m going to list out some potential users/personals

Individual person working from home. During the work week he goes out or orders food for lunch.
Same individual as above but with diet/allegory restrictions.
Family of 4 or more.
Someone that loves to eat out but always wants to try something new.
Someone that loves to eat out but is looking to stick to what they like

So now that we have the above, I would like to prioritize them with the following attributes

MS: Market Size

MG: Market Gap - Perhaps these group of people are underserved in the current search market

I’ll use the 1,2,3 rating. A 1 signifying a lower market size/gap and 3 signifying a greater market size/gap.

	Individual	Allergies	Family	New Food	Same Food
Market Size	3	1	3	2	2
Market Gap	1	3	2	2	2
Total	4	4	5	4	4

So for this exercise I would like to focus on a family (4 or more) as the user personal.

Makes sense.

Next I want to identify some pain points for the person, to see where we can focus on improving search for this user.

I’ll do this by going through the user journey of searching for a restaurant for a family of 4 or more.

Family needs to decide if they will be cooking, dining out, ordering take out, or doing delivery.
Family needs to decide on what food it wants to eat. Japanese, Mexican, Italian, Indian, etc...
Once family makes s decision they need to look for a restaurant. They will search for a restaurant based on several factors
1. How far is the restaurant? Is there traffic going to and from the restaurant?
2. Does the restaurant offer delivery? If so from which service (Door Dash, Grub Hub, etc…)?
3. Does it offer dine in or takeout?
4. What kind of food does it serve?
Family finally selects a restaurant.
Family dines in, gets take out, or gets delivery from restaurant
Food is served/delivery and family dines.

So above I put the typical user journey for a family attempting to eat. Being that we are looking to improve search, I think it would be beneficial to target a pinpoint that the market has yet to fulfill. I think I want to focus on the 2nd point: "Family needs to decide on what food it wants to eat. Japanese, Mexican, Italian, Indian, etc…”. I’ve select this pain point for the following reasons:

If a family of 4 wants to eat, it would be worthwhile to make sure whatever restaurant they pick has food that they will all like.

I haven’t seen any innovate solutions to help people pick a restaurants based on the needs of 4 different individuals.

Now that we have a pinpoint we are solving for, I would like to come up with some solutions.

Search engine where you search by food dish instead of food type (Ex: Looking for a place that has hamburgers and tacos).
VR search engine where you can visually see a prepared dish
Search engine that you an search based on a food pallet

Now that we have 3 possible solution. I want to evaluate some trades offs to see what would be the best solution to implement. Again I will use the 1,2,3 scale.

Take make addition easier, a high level of effort for implementation is 1 and a low level of effort for implementation is 3

	Food Type	VR	Pallete
Level of effort	3	1	2
Innovate	1	3	3
Total	4	4	5

Thus, based on our criteria I would go with the pallet idea.

The way I would implement is as follows: The app would give each individual family a test as to what type of foods they like. Based on the results each of the family would be assign a “pallet” (Spicy food, savior food, etc…). This pallet can then be entered into the search engine and using all the pallets of every family member the search engine will give a set of results.

Interesting idea… can you summarize?

Sure, so in order to improve restaurant search. We decided to focus on families with 4 or more people. We then focused on helping them decide what type of food to eat. From there we came up with the idea of being able to help them develop a food pallet that we can use to search for restaurants. Based on each members food pallet, the restaurant search would return a result that covers all their family member’s food pallet.

**My own feedback**

I just started out practicing for interviews again, so I’m a bit rusty. Thus I welcome feedback on how to improve this answer.

Couple notes on this now that I’ve written it all down.

For the clarification portion, I think it would have been worthwhile if I explicitly also asked which platform we were targeting (web or mobile) and the technology.
In retrospective, I also forget to mention another key player in search: Advertisers.

ajayg · Answer 2 · 2020-12-19T07:14:31+0000

Clarifying Qs/Goal:

What is it we are trying to improve in the restaurant search via Maps: Engagement, Acquisition, Adoption, Retention: Lets say Engagement
Any specific user groups targeted -No
Is it on any specific platform or geo-loc? - Lets say App in US market

User groups: Family with kids , Travelers looking for a quick bite, Those looking for takeout only during covid times

Focus on : Family with Kids -

User Journey

User launches the Restaurants search from Maps.
User can filter by location, type of restaurant, Cuisine, Delivery options, Reservations, Price etc
User will see the individual listings based on the proximity from user selected location.
These listings will have images, Ratings of that place, Duration of operation, Price range
User can Save this choice and Map directions to the place.
User can Reserve a Table

User Pain POints

In A family , each will prefer different meal options especially with kids - Need to have better Menu options
If the restaurant is in a City, it will be a challenge to Park and Pick up food and even to dine in
Sharing the Restaurant info with other family/Friends so they can all meetup there or pickup food from the same place
If in a group of family members, if there are any food / dietary restrictions, how can restaurants provide that info

Prioritized Paint points : #1, #2, #4, #3

Will choose 1, and 2 and 3 for MVP

Priority	Solution	Value/Impact	Effort/Cost
1	#1 - Currently I see Google Maps/Restaurants do not have a real time Menu choosing option. Its only a static image. We will need to have options like 1) kid friendly menu 3) Lunch/Dinner Menu 4) Drinks 5) Desserts .. that easily helps user to pick and choose.. Making ordering easy/fast	High	Medium
2	Solve the parking challenge by 1) Notifying if there is a drive through option 2) Any parking lots/parking spaces in the vicinity- surface those options.	High	High
4	Give more details around the Menu with an eye towards any dietary restriction family can have - Nut allergy, Vegan options, etc	Medium	Medium
3	Have a Share option within each Restaurant - to be able to share with other family members or friends.	Medium	Medium

Metrics:

%DAU, WAU, MAU who use Restaurant search from Maps and access the Menu

%DAU, WAU, MAU who use Restaurant search from Maps to Share it with others

%DAU, WAU MAU who look up Parking Spaces nearby the restaurant

%increase in order volume in a Week/Month at a restaurant when they are searched via Maps

# of reservations in a Week/Month via Maps for restaurants

solders15 · Answer 3 · 2020-04-16T18:19:03+0000

Problem: Improve the restaurant search experience in Google Maps

Me(M): Why and What do you want to improve?

Interviewer (I): Improve engagement/retention. We have found that customers use other services to choose restaurants

M: Is there a reason why other services are being preferred?

I: We are sure, overall the experience seems to be better in other dedicated services

Framework that I will use: First, I will list potential customers, select a segment, list their pain points, select a few of those painpoints, present solutions that help engagement, prioritize them and the measure how the address the goal of engagement based on metrics.

It is to be noted that in this question, there is no issue of Awareness, Acquistion and Activation. This is because of the dominating market share of Google and the widespread knowledge of what Google maps does. Is that a reasonable assumption?

I: Yes.

M: Here are the two main customers and their pain points

Restaurants (i.e. their owners) – more the choices tailored to Google Maps better the retention

- Need to understand what the user is looking for

- Display items that pique the user’s interest (cost, waiting time, ratings, photos of food choices based on user signals)

- Best way to go to the top of the displayed list.

- Ability to offer specialized coupons to users

End users: People who want to go to a restaurant alone or in a group.

For the purpose of this interview I’m going to choose the end users as they drive the engagement/retention

Journey – find restaurant -> go to restaurant -> eat at restaurant -> pay – get back to home/office or destination

(1) Difficult to choose the restaurant, -- many choices

(2) Coordinating with multiple people, -- linking with calendars

(3) Communicating arrival time with restaurants so that they can plan service better.

(4) Total time of meal (travel + eating) – so that everyone’s schedules can be met.

(5) Optimize cost of entire meal. – deliver coupons/ indicate the approximate cost for the entire party

Based on the goals – engagement and retention, I will choose (1) and (3) . Within (3) I would like to focus on optimizing the that provides the best differentiation against other products and will drive people to use Google maps as their restaurant scheduler.

Solutions (Impact, Technical, time of implementation)

Pain point 1: Difficult to choose the restaurant

Use signals from the user and the restaurant owners to display an ordered list of 5 restaurants

Signals could include one or more of (Cost, distance, quality of food (reviews), type of cuisine, waiting time). Signals could be gathered using past history of usage or an user profile. (H, L, L)

If multiple users are involved, then a combination of signals from all the users could be used to decide the ordered list – (H, M, L) - This could also mean sending a message to the other users with a list of restaurants and their associated characteristics and gather information and then finalize the list and display to everyone.

Display restaurants that allow you to order remotely via an interface and keep food ready to go based on when you and your friends get there

Pain point 2: Arrival time communication with the restaurants

Based on current location of set of friends, and knowledge of their transportation status, Maps suggest the best way to get to the restaurant so that everyone arrives around the same time – (H, H, H).

This would potentially involve sending an Uber at the appropriate time.

Communicate the time of arrival of the entire party to the restaurant so that they can plan the service better.

Message can be sent to an email/SMS/Hangouts chat

Prioritization

Based on the list of solutions I would choose the “tailored” ordered list to the user as well as communicating the arrival status to the restaurants to the restaurants.

Metrics

Increase in the number of users using restaurant search (A/B test if needed)

Increase in the number of restaurants providing specialized services (coupons)

Increase in efficiency metrics using feedback from restaurant partners

Potential revenue from transportation partners.

Number of return users/week/month

Summary

Based on the goal requirement, I analyzed the different customer segments, chose the segment that will drive engagement and presented solutions that will be a differentiator for google maps as compared to the competition so that engagement/retention goals can be met.

productmanager123 · Answer 4 · 2021-06-21T02:07:24+0000

Clarifications

Current product experience: I can search on Google maps mobile app or webapp for “restaurants “, and Google maps returns locations which are relevant to that search. And these locations can be restaurants or not. I usually use only the search bar and nothing more. So I dont know if there is any restaurant specific search experience in other screens. So the UX, I think is currently available is - and here I am focusing on Google maps mobile apps. Google search web app - thats also lists restaurants - but I am assuming thats out of scope.

Lets look athe mobile app exp:

1/ I can search on the search bar.

2/ I get a list of results of restaurants: TYpically what I have seen. If its generic search such as “resturants, or indian restaurants etc”, I get a list of search resutls which are essentially “near me” in some sense. I think google uses, distanc efrom my current location as one of the factors in the ranking algo in addition to match to key word, reviews, ratings etc.

3/ The search results show up either is lsit view or I can see them in the map itself with pins. I think I can add filters “open now”, “what is my budget etc”, “avg rating”, and teh results are filtered out. I dont think there is a sort.

4/ I can browse the search results by scrolling - basic info is listed on the seach result cards -- # of reviews, Avg rating, location address, how far it is from my current location, thumbnail photo.

5/ When I clock on it, card blows up to show -- link to website, call to talk, address info, a few reviews which i can scroll and see more, link to menu etc.

6/ I think I can also leave a review, though I never dotn taht - so dont know if it can be doen in the app or need a different app.

What I cannot do: Reserve a table, order take out. Basically cant do any trasaciton.

The other side of the equation is restaurants - I am assuming that theri information is automatically indexed form their website. Restaurants can go ahead and edit teh google index, add more information as needed. There must be a process for arbortating on unfair review etcs.

So improve restaurant search, I first want to understand Google Maps strategic objectives:

1/ Acquire new users

2/ Increase engagement with existing users

3/ Drive up revenue

Do you have any feedback on which of these goals are important for Google right now? No.

OK. I will priroitze increasing engagement with existing users. If you drive up engagement, you are inherently creating more value, which will automatically drive new users to come. If there are lot of users and engagement is high, then revenue will ake care of itself.

And I am prioritizing end users here. Again, if I can aggregate demand from end uers, then restaurant wil automatically follow.

Before I get started, is teher any specific reason, you want me to focus on restaurant searhc? And where does this scoe end? Is it to provide the best first screen results to the user - when they click, the subsequent screens - do thye fall under search purview? Yes. Everything is under seach, till the user bandons the search intent and does something else.

To increase engagement, lets look at the problems user faces during the customer journey we discussed previously. If I have more time, I would also look at competition - there maybe a lot of table stakes that comeptiros aledy provide. For this discussions, I am keeping competition analysis out of scope.

Pain point	priority	Solution	Cost and
Awareness of whats possible: I dont know that I can search for restaurants on Google maps. I mean I know that I can search for specific locations such as KFC, but I dotn know if I can search for “mexcan restaurants” which is a more broad.	M-H. I think this is a big problem. Many time even I dont know that the feature is possible until I actually try it.	Show “mexican restaurants in Sammamish” as a search shortcut on the main screen. Basically based on the user restaurant search history, provide a pre-built search keyword. If not user history on restaurants, then show most popular search keyword for the location and based on other attributes of user profile	Landing screen is high value real estate. Implementation: L-M. Should A/B test this. Test this on users who didnt do abstract searches in last month, but who searched more than once per week. Metrics: Number of abstract searches/ Number of searches per week/ per user. Will also look at overall Google Maps key metrics. THis should not -vely impact main metrics (DAUs, searches etc) for Maps overall.
Ok. I decided to searhc. Now help me make a decision. Typically the decisions I have in mind are a/ I have a specific cuisine in mind. Show me something which I would like and fits in mt budget and is not too far away (getting relevant and raking right) b/ I dont know what I want, I just want to eat outside -- too vague. In both cases, help determine which restaurant to go at a high level. c/ Once I knwo that I want to go mexican, then which restaurant? I want to go a place which is affordable, which has high reviews and ratings, which I can trust is a good place to eat. d/ Once I decie which restaurant - 1/ I want go there or do a takeout or get delivered. Go there - reserve a table, dont have to worry about parking, and would like to be seated as soon as I arrive, 2/ or I want to place an order, and pick it up 3. Or place an roder and get it delivered.	H. Finding the right restaurant is a lot of work. H Reserving, qordering the menu before we arrive. Placing order for take out, delivery	Reviews and ratings help, but if we can social proof thats most valuable. Your friends liked this restaurant. Basically use social proof as key attributes in ranking.	Very HIgh difficulty. Google doesnt have social network. One way to connect - use phone network. If there is a contact in your phone network, and has left a review, then surface that. Its low signal, but still better. Also this will help in inferring the social network and its relationships of the person without explicitly asking for it. There maybe privacy issues here. Need to think through. Reserving a table and putting in order. M. Needs integration with restaurant systems. But resruatrnats from incentive to do this. They will integrate with Google API. Need to drive awareness and drive integration. Placing an order - once you have menu, then you can order for takeout. M Getting it delivered - will need to setup fulfillemtn service like Doordash. VH
The entire above experience - how can you make it completely voice based? My mom finds it difficult to use the screens one by one. But dont want it voice in the sense of Alexa - it should be voice + sight.	M Not sure if this is a big problem		H.

In summary, will focus on two features (a) driivng awareness through “pre-built” search, (b) enabling to reserve table and place an order before reaching.

KN · Answer 5 · 2020-04-26T12:03:15+0000

Okay. So i will approach the question in the following way:

I will first identify the Problem at hand. What are we trying to do? What is the Goal of the product improvement?

Second, i will identify the users for whom we are trying to imrpove the Product for.

Third, i will see what are the unsolved user problems and what are the gaps and improvement areas.

Fourth, i will suggest the improved product layout.

Fifth, i will provide product success and validation metrics.

Sixth, i will summarise my recommendation.

Step One: What is the Problem at hand?

The problem is to improve restaurant search.

Why to improve the search? What is the goal? Is it revenue or engagement?

I will take engagement as the goal. Idea is to build a product that users will make their first choice in food search and then later monetise using Ads or some other mechanism.

Step Two: Identify the User

The Users can be segmented into two categories:

Dine Out Users:

They want to go to the restaurant and have a meal with friends, families or by themselves. Restauarnt can be a fine dine, fast food, or tea/coffee house etc.

They are on the lookout for places that are near to them and offer good quality food. Price may or may not be a factor for them.

They are generally looking for places that have good reviews and recommendations.

Home/Office Delivery Users:

They want their food to be delivered at their home/office. Restaurants are generally fast food, tea/coffee houses or speciality food restaurants like Indian, Continental, Chinese etc.

They want quick delivery of food. Generally don't like to spend too much on home delivered food. Are not very brand conscious as they are looking for quick ways to get a meal.

Steps Three and Four: Coming over to the unsolved user problems by Google restuarant search and the improvement areas. We can refer the below table for the same:

User	Problem	Solved or Not	Difficulty	Priority
Dine Out/Delivery User	Search for available restaurants	Solved
Dine Out/Delivery User	Filter based on ratings	Solved
Dine Out/Delivery User	Open Now/ Open at X Hours	Solved
Dine Out User	Visited in the past	Solved
Dine Out/Delivery User	Cuisine	Solved
Dine Out/Delivery User	Sort by Relevance/Personal match	Solved
Dine Out User	Occupancy Info	Not Solved	Medium	3
Dine Out/Delivery User	Dish Based filter	Not Solved	Easy	1
Dine Out/Delivery User	Dish Based Speciality	Not Solved	Medium	2
Delivery User	Filter/Sort on Delivery time(approx..)	Not Solved	Easy	4
PickUp User	Set Restaurant Marker on Maps for a multi way trip	Not Solved	Easy	5
Dine Out/Delivery User	Voice Based Search	Can be improved	Difficult	7
Dine Out User	Friends eating or going to a restaurant	Not Solved	Easy	6

I can show the occupancy rate basis crowdsourced Maps data and approximation algorithms.

This will help me determine if I should go there or try booking a seat there even without trying to call there.

There can be a dish based filter. This can be implemented by parsing the menus of the restuarants.

Also, the Reviews can be analysed using sentiment analysis to determine a Speciality score for major dishes at a restaurant.

I can also have a sort/filer feature based upon the delivery time shared by the restaurants. However, this will be approximate data and it should be combined with distance, traffic conditions and restaurant occupancy data to highlight any large changes.

For the User who picks up his food, I should have a multi way trip option where I start the trip, pause it using GA and re start using GA again to reach home.

Voice Based Search can be improved to include filter and sort options in it.

Finally, if I want to know where are my friends/contacts eating now or the next 2/4/X hours, I can see that so that I may organise a get together with them.

Step Five: To measure the impact of these, we will have to measure the engagement of the users before and after these feature launch.

Time spent on Restaurant Search
Number of Users doing Restaurant Search
Number of Users using new features
Frequency of Usage of New features and old features

Step Six: Recommeded to go with Feature with Priority 1-6, Priority No. 7 is difficult to build and offers little improvement over current product other than adding better voice capabilities.

Increased engagement wih users and restaurants will lead to better revenues in the food tech space.