How would you improve Yelp?

Question

How would you improve Yelp?

Asked at Walmart over a year ago

3.2k views

Product Improvement category

Asked at

3.2k views

How to answer Product Improvement questions

Interview Guide

Liked by PM Exercises

P

Gold PM

Jul 4, 2020

Interesting. So I assume I'm a PM at Yelp and my goal is to improve the product.

In terms of approach, I would like to align on what Yelp does for their users, identify potential goals they are trying to achieve along with specific metrics, then go deeper on user types and use cases, prioritize those use cases, and finally make a recommendation. Does that work for you?

Ok, so I see Yelp as having two main types of users, businesses and consumers. For businesses, it gives them the exposure to a huge consumer market, and also provides them with the ability to get feedback to improve their operations. For consumers, it works as a discovery engine e.g. if I want to try a new restaurant I have not visited before, I can filter by cuisine, location, and price. It also helps users make decisions on which businesses to visit through transparency around overall ratings and specific reviews. Finally, it also allows consumers to take action such as booking a reservation at a restaurant.

As the mission, I see yelp as a product that helps consumers discover great experiences. And they do this by matching consumers with the best businesses based on their needs and interests. In terms of their main product goals, they could be related to acquisition, engagement, or monetization. I think the most meaningful would be increasing consumer engagement given that:

Yelp is already a mature business, and that if you get enough engagement, revenue will follow.
If you get enough consumer engagement, business engagement will follow.

In terms of metrics, I would consider # of Weekly Active Consumers as a potential great metric to start with. Does that make sense? would you like to explore different metrics? No? Great, let's keep going.

Ok, now that we defined the goal, let's think about user personas within the consumer user. Here are a few that come to mind:

Foodies. Core use cases are around finding new restaurants to try out. Mostly urban and trending younger.
Power reviewers. Core use cases are around trying new businesses and producing valuable information to share with other consumers thinking about visiting or using the same business. They care about building a strong online reputation for the quality of their comments.
Home owners. Core use cases are around finding service providers to help with home improvement or fix any issues e.g. plumber, landscaping services, electrician.
Travelers. Core use cases are around discovering things to do in places they visit. They often are interested in doing less touristy things and getting the inside scoop from locals.
Parents. Finding schools and daycare centers, playgrounds, sports, activities, camps for the kids, plan birthday parties. Very seasonal in nature.
Other.

Let's prioritize and focus on one. Foodies and Power reviewers are large segments that helped build Yelp, so I believe they're super well addressed. Home owners probably are less frequent users, but the product seems to address their issues really well. Travelers are a super interesting segment, but at this point I don't think we would be able to make an impact on the metric we are trying to drive due to the uncertainty related to the pandemic. Parents are super interesting in that I assume they're not as engaged with Yelp - huge upside - due to the lack of product features to address their needs.

Let's think about a few use cases for parents and then let's prioritize:

Finding camp(s). Parents with kids from 5 to 17 need to find options for their kids to learn and grow during school breaks. These could be sports oriented or general, and could be long (6-8 weeks) or short (1-2 weeks).
Planning birthday parties. Every year parents need to plan birthday parties for their kids. Whether you choose to do it at home or at a park, there are tons of logistics related to making it happen, which are hard to manage on top of all other parent commitments.
Helping find sports/activities. These are usually very seasonal and require parents to be on top of what is available around them to help their kids find their passion. Mostly targeted at youger kids. After kids discover their passion they usually stick with it for middle school and high school.
Finding the best schools for their kids. Only happens relatively infrequently.

Use Cases	Reach	Impact	Confidence	Effort
Find camps	Medium	Medium (a few times per year)	Medium	High
Birthday party planning	High	Medium (on average 2x per year, but highly involved)	High	Low/Medium
Find sports/activities	Medium	Low (more important in elementary school)	Medium	Medium
Find schools	Low	Low (not that frequent)	Medium	Medium

I would focus on "birthday party planning" due to it's potential to reach every parent and due to its relatively low effort given that most businesses required to plan a birthday party are already on Yelp. In terms of solutions, I would consider:

Creating a landing screen for planning birthdays
1. Broken down by age, given that the experiences are so different depending on the age group e.g. teenagers would probably prefer to go to a movie with friends.
2. Start with semi-manually curated plans (based on review analysis):
  1. At-home plans - getting meals, entertainment.
  2. Outside of home plans - e.g. park, playground center.
3. Allow consumers to create plans - e.g. kara's cupcakes + domino's pizza + ABC wine shop for adults - with a checklist to manage them.
4. Allow other consumers to start from an existing plan and tailoring it to their needs.
5. Allow consumers to vote on the plans.
6. Allow consumers to create invites for the birthday parties and manage attendance (powered by Yelp).
7. Allow consumers to order birthday items from businesses.

In terms of MVP, in order to achieve our goal of improving # of Weekly Active Consumers, it would be important to include the personalization of the plans (so that parents can come back to it in the weeks prior to the birthday event). This would include basic versions of 1.1 through 1.5. Later on we could add additional functionality to integrate with down funnel (1.5 and 1.6). I would work with Eng to estimate effort, and with Data Sciences and the Editorial team to come up with starting points for "template plans".

3 likes | 1 feedback

1 Feedback

Bijan

Platinum PM

Jul 4, 2020

Things you did well

Structure: Great structure of the answer. It's easy to follow and see that you are familiar with answering product design questions
User groups: You broken down the user group into mutiple groups and picked a particular one. I really like how you went about describing each ouser group and then explaining why you'd like to focus on parents.
Pain points: You listed a good number of meaningful pain points / user needs
Solutions: Great set of solutions to solve for the pain points listed earlier
Evaluation of features: You compared the solutions based on meaningful criteria
Out of the box ideas: Good list of solutions, which speaks to your ability to brainstorm various ideas

Areas of Improvement

Describe tradeoffs: If there is time, describe what sort of trade offs your suggested product will have. Are there any risks you have to keep in mind when building this product?

Great work overall. Really enjoyed reading this answer!

0

Top Walmart interview questions

How many dentists are there in New York?33 answers | 36.7k views
How do you decide how many cash registers you need for a new Walmart store?33 answers | 39.4k views
Estimate the number of Uber rides.30 answers | 51.3k views
See Walmart PM Interview Questions

Top Product Improvement interview questions

What is your favorite product? Why?89 answers | 263k views
How would you improve Google Maps?53 answers | 228k views
How would you improve YouTube?29 answers | 81.3k views
See Product Improvement PM Interview Questions

Invite members
Invite by email

Add another

Rafa · Answer 1 · 2021-09-27T07:54:47+0000

How would you improve Yelp?

Additional information and clarification questions

Since Covid19, there has been a drop in engagement around restaurants that we would like to address.
The drop is in # of orders, # of screens seen per session, and # of reviews given.
Can we focus on the mobile app in the US? (Yes)

Structure

Mission -> Product -> User Segments -> Pain points-> Initial product -> Metrics and Concerns -> Summary

Mission and product

Yelp’s mission is to connect people with great local businesses. The product allows searching restaurants, reviewing menus, reading and leaving feedback, and order sits in the restaurant or order dishes. AFAIK, Yelp isn’t responsible for the deliveries.

User Segments

Restaurant owners.
Users
1. Power users – use the app almost Every day.
2. Regular users - use the app several times a month.
3. Occasional users - installed the app installed but almost don’t use it.

I would like to focus on users that struggle to use the app during the pandemic.

Many of the problems related to covid19 are common to many user segments, yet I would like to narrow the scope to users who use the app often, power and regular users.

Pain points

Reviews relevance - most of the reviews left pre-covid19 might not be relevant since the experience changed significantly.
Restaurants data relevance
1. Is it open?
2. Can we order a table?
3. Do you have all the dishes that are on the menu?
4. Are you doing deliveries?
5. How long should it take?
I would like to order something for a friend who is in quarantine.

Prioritize pain points

	Impact	Size (potential)	Reach (%)
Reviews relevance	High *1	High	High
Restaurants data relevance	Low *2	Medium	Medium
Order for someone	Very High	Medium *3	Medium-High

*1 The primary value proposition of Yelp
*2 Some people will call the restaurant to verify the information, and if something is not accurate, they will leave a comment in the reviews\comments section.
*3 Although it is trendy, most people won’t order deliveries for others.

Therefore I would like to focus on #1 - Reviews relevance.

Solutions

Yelp will classify reviews manually as being given during covid19, affecting the ranking (it can’t be done automatically since not all people leave reviews right away).
Yelp will classify reviews using machine learning as being given during covid19, affecting the ranking.
Allow users to classify feedback as “during covid19”.
Add a widget or a section for top orders during covid19.

Prioritize solutions

	Impact	Effort
Manual classification by Yelp	High	Low-Medium
ML classification by Yelp	Medium	High
Classification by users	Medium-High	Low-Medium
Covid19 widget \ section	High	Medium

#3 is better than #1 since users can classify items much more accurately and faster than an admin.

I would like to focus on #4 and #3; it will improve the app’s relevance for the pandemic and data accuracy.

The initial product

Use the current widget of top recommendations around you.
Add a UI element in the table to let users know that data is up to date for Covid19.
When leaving feedback, add an option to select whether the review is given based on experience during the covid19 (set by default).

Metrics

Primary Metrics

Clicks / Exposure (# unique users).
# of users engaging with the widget.

Concerns and counter metrics

Demoting popular places - can affect Yelp’s relationship with them.
# reviews to other sites.
Impact of Bots \ fake reviews is much higher if start from scratch.

Summary

We would like to improve the engagement of the Yelp mobile app for US users during the Covid19 pandemic by adjusting the app to be more relevant and accurate for the situation.

We will do that by classifying reviews as being given during the pandemic and update the recommendations section to give heavier weight to the latest reviews. Since Yelp has many users, the effectiveness of those changes can affect pretty quickly.