How would you improve YouTube?

2. Clarification questions
 

1. What do you want to improve in the app? And why? Revenue/ Engagement/ Retention/ Adoption? -  Say revenue. This is because YT has a user base of more than 2.5 MAU but the paid subscribers are around 3.2%

2. Shall we also consider how much are we planning to increase? Say 1%

3. Is there a specific channel to be considered (Web/Mobile/ Tablets)? No, overall

4. Do I own the entire app or am I the PM for a specific vertical? Say the entire app

5. Is there a geography that we are looking at? Say India

6. Is there a timeline that we are looking at? Say 6 months

7. Shall we improve an existing feature or build a new one? Anything

3. Choose a metric to be improved - North Star metric under the umbrella of the main objective - Revenue. We have to increase the number of people who buy the YT subscription

The user group that currently has YT premium in India is the urban youth with some extra money to shell out. The main reasons they buy it are ad-free content and a background player for music or podcasts. Let us try to tap into some other personas. 

4. List the user groups and select ICP 

Multiple user groups can exist like

• People who like to learn on YT

• Health enthusiasts - Ad-free content can help them

• Parents who have YouTube kids - Ads can be disruptive to studies

• Elderly people who watch videos for recipes or home exercises but get confused with ads

5. ICP 

• Parents who have kids below the age of 10

• Urban and willing to shell out the extra bucks if they see value. Not highly price-sensitive

• Likes to give screen time to their kids

• Wants to control and monitor what their kids watch online

5. Pain points of the selected user group

• No control over what the kids watch within the app. Some animated kids' content also has violence and nudity.

• Does not have the time to go through the watch history every single day to monitor

• Has a busy schedule so they have to give screen time to their kids

5. User journey 

The user buys YT premium and gets YT music and YT kids for free. Both these apps are optional but come as a package with a premium subscription 

6. Solutions 

In YT premium let us package some functionalities of YT kids

1. Set a profile for kids based on their age and include the topics they want their kids to see on a broad level

2. Show only those topic videos to the kids

3. Send a weekly report on what the kid watched in the week and point out if there are any out liars from the watch history

4. Should be able to set the timer for when can the app be switched on and when it needs to be switched off to control the duration that they watch

5. Prompt to update the preferences from time to time to keep the kid engaged

7. Evaluate solutions 

8. Define metrics for measuring performance - These metrics are wrt YT kids post purchasing premium 

Acquisition 

• # of YT kids app download

• # of YT kids signed up

Activation/ Engagement

• Active Users: Daily Active Users (DAU), Weekly Active Users (WAU), Monthly Active Users (MAU).

• Session Length: Average duration of user sessions.

• # of parents who customized the view

• # of parents who used the setting time duration feature

• # of parents who view the weekly report 

Monetization 

• Average Revenue Per User (ARPU): Revenue generated per user.

• Customer Lifetime Value (CLTV): Total revenue expected from a customer over their lifetime.

• Churn Rate: Percentage of customers who stop using the product within a given time period.

• Overall revenue

Retention

• # of users who come back after the period (daily, alternate days, multiple times a day, etc.)

Satisfaction 

• Net Promoter Score (NPS): Measures user satisfaction and loyalty by asking how likely they are to recommend the product.

• Customer Satisfaction Score (CSAT): Measures user satisfaction with a specific interaction or overall experience

YouTube is a media platform that allows users to discover various media assets created by creators. It sits well with the Google ecosystem because Google is all about organizing information. So, gmail is organizing your emails, YouTube is about organizing the content that is uploaded and personalizing those recommendations to the viewers based on their knowledge graph. YouTube also has many parts to it. It has YouTube, YouTube Music, YouTube Kids. In the US, it also has YouTube TV that allows people to watch TV from various networks.

I have a couple of clarifying questions. 

1. Is there any specific part of YouTube we are focussing on, such as YouTube Music, YouTube Kids or YouTube the main video product? 

2. Also when you say improve, are we looking at specific improvement areas, such as acquiring new users or engaging the existing users? Do we have any specific data points that we should look at to improve?

3. Additionally, are we looking to improve any specific channel, such as mobile or laptops?

I will look at the main YouTube channel with the objective of retaining users, given that YouTube is now a more mature platform. Also I will look at mobile as the channel, because increasingly that is the mode of video consumption.

One final question. YouTube is a 2-sided marketplace of sorts, with the content creators and the consumers being the 2 sides. There is a flywheel in effect here because the more the creators create, the more the consumers will consume, and this will lead to more videos being created. Is there a particular side you want me to pick, or can I just pick one side. 

I will pick the consumer side, because engagement at that level will also help the flywheel and encourage the creators to create more videos. 

So, just to confirm, I will look to improve engagement on the YouTube main channel, and for the purposes here, I will look at the content consumer side. 

There are primarily 2 groups of people who come here - Knowledge seekers and entertainment seekers. I will look at entertainment seekers for this particular question because there are many other distractions for entertainment seekers, such as Reels, TikTok, Netflix, Prime etc. that it would be good to increase the average engagement time of the users who come to YouTube to seek entertainment.

So, the 2 pain points that I want to focus on are : 

1. Unable to find a channel that he has been following

2. Where the user has limited time (say 15 mins) to watch a longer video

    

    

   Now, let me list solutions for each of these

    

   Pain Point	Solution	Impact on Engagement	Engineering Effort
   User is unable to find a channel he has been following for long	If YouTube has taken down a channel for violation (such as copyright violation), then all the subscribers and users should be immediately notified as such.	M (This should already be happening but I didn’t see it when one of the channels I was subscribed to was taken down for copyright violation)	S to None. Just a notification that needs to be sent.
   	When the user searches for the channel, YouTube can give information saying that discussions are on with the creator, and in the meantime, you can watch videos from the other creators who are in the same space.	L (This will ensure the bad taste is not there and the user is still presented content that can keep him engaged on the platform)	S (Not much effort is needed as the same knowledge graph can be leveraged)
   	Alternatively, a new channel can be created for the creator (especially if it is a minor or first time offence) and the subscribers can be migrated to the new channel, and the consumers can be notified of this. This new channel will have none of the old content but you are just giving a chance to the creator to continue his work until he fights his case with YouTube.	M(Not much impact on engagement. You might have to ask them to subscribe to the new channel and this might even reduce engagement as they may consider it spam)	M (migration needs to be built in. Also, we will need to consider legal and other teams before doing this.)
   User has limited time to watch a longer video	If the user can only watch 15 mins, and then comes back a couple of days later to resume, then you can give him a quick recap of what he saw the last time	L (Most users will remember a brief of what they saw, so it will not be a deal breaker)	S (Gemini should be able to do it relatively quickly)
   	Alternatively, you can create a custom TV schedule for the user based on his free time. We can ask the user to block his calendar with the word YouTube from 8-9 PM for eg. We can then create a TV schedule based on what he wants to see/continue. He can keep adding things to his saved videos and we will create a weekly program calendar for him based on his preferences.	L (Building a custom programme schedule in YouTube will definitely help users remain engaged, and it will also increase premium subscriptions if you want to do this on Premium, or ad subscriptions if you want to do this on the free tier. This is also one way we can take on the streaming behemoths,  especially given the variety of content across industries like sports, movies, podcasts, etc.)	L (This will take some effort because we need to take into account the free time of the user, preferably pulling that in from the calendar and then checking the time of the podcasts etc. We also need to take the interests of the user, and also allow him some time to discover new content)

Based on the above analysis, in the short term, I will look at letting users know when the channel is taken down and give them recommendations on similar channels to follow. Additionally, to solve the problem of users seeing contents in episodes, it will try to put in some smart summaries and clips to just give a recap of the last part that they saw so that they can continue seeing the content. In the long term, it would be good to build a personalised program schedule based on the user’s interests and watch history along with the time he wants to watch YouTube.

The metrics I would look at when I build these solutions include

1.  Amount of time spent on YouTube 

2. Number of new channels subscribed per user

3. User churn for channels taken down (Of the subscribers that were available, how many churned and how many were engaged with other channels)

    

    

   To summarize, we picked YouTube for consumers on mobile and we picked engagement as the goal. We identified a few pain points, of which we picked these 2.

   1. Unable to find a channel that he has been following

   2. Where the user has limited time (say 15 mins) to watch a longer video

    

   For each of these we brainstormed solutions, and found that in the short term we can use the following solutions.

   1. When the user searches for the channel, YouTube can give information saying that discussions are on with the creator, and in the meantime, you can watch videos from the other creators who are in the same space.

   2. If the user can only watch 15 mins, and then comes back a couple of days later to resume, then you can give him a quick recap of what he saw the last time

    

   In the long term, we will plan to give a personalised programme based on how much time they have to spare and we will customize the programmes accordingly.

    

To address the question of how to improve YouTube, we can establish a set of assumptions and concepts to tackle this broad and ambiguous issue. The first concept I introduce is the Product Funnel, which consists of three stages:

1. Promise Stage: The marketing team promises the user that by entering the app or website, they can solve their problem.

2. Need Expression Stage: The user accepts the promise and enters the platform to express their need in some way.

3. Need Fulfillment Stage: As a product, I strive to satisfy the need that the user has expressed.

Since the Product Funnel concept is applicable to all products, it can be defined for YouTube as well:

1. Promise Stage: YouTube makes various promises to users through advertisements on Google Play, and other platforms, or by showing video previews in Google search results.

2. Need Expression Stage: YouTube offers different ways for users to express their needs, such as allowing them to perform searches, explore to find desired content, or subscribe to channels.

3. Need Fulfillment Stage: After the user has expressed their need, YouTube provides solutions, such as showing videos, possibly shorts similar to Instagram Reels, or displaying channels.

However, we all know that YouTube does not make just one promise; it offers many promises to different personas. Therefore, I introduce another concept called Use Case. YouTube has numerous use cases, some of which include:

• Hobby Seekers

• General Information Seekers

• News

• Sports

• Kids

• Specialized Content Seekers

I have introduced two concepts so far: Product Funnel and Use Case, and I have tried to describe my product up to this point. Returning to the question, "How would you improve YouTube?" it seems very general, and as mentioned earlier, no specific metric has been provided to us to solve the issue. Therefore, I take a little inspiration from the divide and conquer algorithm. I think if we can break the problem down into smaller parts and then solve each part separately, we can ultimately claim to have solved the entire problem.

To improve YouTube, I would focus on enhancing the Product Funnel and Use Cases. Here is a detailed approach:

Divide One: Product Funnel Enhancement

1. First Stage - Promise: To improve YouTube at this stage means to make a better promise to the user. This is primarily a marketing task, but a Product Manager, especially one with a Marketing Product Manager focus, also plays a role in improving this funnel.

2. Second Stage - Need Expression: Improving YouTube here involves finding a better way for users to express their needs. This could mean creating a better UX Funnel, which is mainly the responsibility of product designers, although Product Managers are also involved.

3. Third Stage - Need Fulfillment: This is where a Product Manager can shine by finding better ways to satisfy user needs.

Based on these stages, I would choose to focus on the third stage of the funnel to solve the problem, aiming to provide a better way to fulfill user needs.

Divide Two: Use Case Specification

For the Use Case, I would select Specialized Content Seekers—individuals who know what they are looking for, such as specialized content. For example, those wanting to learn C# or seeking tutorials on User Story Mapping. This persona resonates with me as I use YouTube for learning specialized content.

However, these users don't only use YouTube for their needs; they might turn to other platforms offering specialized content. For instance:

• Coursera: Allows users to set goals for courses and provides reminders.

• Quera: Offers Q&A forums or discussion forums.

• Udemy: Provides the opportunity to go through courses in a project-based manner.

Divide Three: Competitive Advantage and Interactive Content

Improving a platform, according to business management experts like Ash Maurya or the lean thinking of Eric Ries, means being better than the alternatives so users don't need to go elsewhere, which is the essence of Product Market Fit. If I can clone the advantages of competing platforms into YouTube, I can claim to have improved it overall.

Conquer:

Another idea for a better way to satisfy user needs is interactive content, where videos ask questions, and the viewer's responses determine the video's path. To measure success, I would run an A/B test, showing the interactive content feature to some users and not to others, and then calculate the engagement levels.

This systematic approach defines the product with the concepts of Product Funnel and Use Case, then uses divide and conquer to break down and solve the improvement challenge, ultimately establishing metrics to gauge success.

This is the framework that I'm going to use to answer this question. In the actual interview I would not mention the framework intitally, but rather dive right into the first part (Product Overview).
 

1. Product Overview
2. Clarifying Questions
3. User Personas
4. Pain Points For One Persona
5. Solution Brainstorm for One Pain Point
6. Prioritization of Solutions (ICE score)
7. Metrics to Measure Success of the Top Priority Solution

1. Is there a specific goal for this product improvement? 
2. Is there a specific metric which Youtube is trying to improve?
3. Are there any other constraints to consider here (for example timeline)?

1.  Viewers:
   1. Regular viewers (viewers who are subscribed to one or more youtube channels and regulary return to watch videos from their subscriptions).
   2. Casual viewers (viewers who don't have any active subscriptions, and watch random videos from youtube they find on other websites or browser youtube occasionally to see if their is anything interesting to them).
2. Content Creators:
   1. Professional content creators (content creators who are creating videos as a primary job, whether its monetized through ad revenue or from some other model such as merchandise sales or patreon subscriptions.
   2. Aspiring profesional content creators (these users may have different pain points than users who are already creating content as a primary job).
   3. Casual creators (users who upload videos or stream occasionally).
   4. Users who upload videos to share with family or friends (could overlap with casual creators).

1. Monetization - these users may be having trouble figuring out how to monetize thier content creation. Maybe the advertising revenue they are generating from videos isn't sufficient, or they types of videos they are uploading aren't eligible for monetization from advertising.
2. Poor presentation skills - some user may have trouble presenting on camera, or speaking in front of a live audience.
3. Problems growing their audience - some users may be having issues attracting a larger audience for their videos.

1. Develop an AI feature that makes it appear as if the user is always looking at the camera when speaking on a youtube livestream. There are other apps which are doing this right now for streamers, but as far as I know their is not a solution for this built directly into Youtube.
2. Develop a Youtube creator academy which has online courses to teach aspiring professional content creators about different aspects of content creation.
3. Offer personalized one-on-one creator coaching services. This service could pair successful creators with aspiring creators.

1. For the first solution:
   1. Impact = medium (I'm not sure how big of an issues not making eye contact with the camera really is for this persona).
   2. Confidence = low (I would need some data or qualitative feedback to give this a higher confidence rating).
   3. Effort = high (as with any AI solution it will take awhile to train and develop the model).
2. For the second solution:
   1. Impact = high (I could see this being really useful for users as it would cover more aspects of presentation skills than just making eye contact with the camera)
   2. Confidence = medium (I would need some data or qualitative feedback to give this a higher confidence rating).
   3. Effort = medium (There are some unanswered questions here around how the monetization would work for this solution or what type of content would be covered. The scope could be very large depending on how this feature is planned).
3. For the third solution:
   1. Impact = high (I think getting direct feedback from someone who is successful in the field you want to be successful in is invaluable).
   2. Confidence = medium (I would need some data or qualitative feedback to give this a higher confidence rating).
   3. Effort = medium (This could be as simple as pairing people for zoom calls, or as complicated as building a dedicated area in the youtube app for this type of interaction).

1. Feature adoption:
   1. Percentage of content creator users accessing the creator academy feature within 7 days or 30 days of account creation.
   2. Percentage of content creator users completing the creator academy feature within 7 days or 30 days of account creation.
   3. Funnel conversion rate from starting the creator academy to completing it.
   4. Total number of users who completed the academy per day/week/month.
2. Retention:
   1. Content creator user retention (measured however youtube is currrently measuring it) segmented by users who have completed the creator academy vs users who have not completed it (or never accessed it). This would allow us to see if there is any correlation between the use of this feature and retention.
3. Engagement:
   1. Average number of session per content creator user per day/week/month (segmented by users who have completed the creator academy vs users who have not completed it (or never accessed it).
   2. Average time per session per content creator user per day/week/month (segmented by users who have completed the creator academy vs users who have not completed it (or never accessed it).
4. Metrics for videos:
   1. Average number of likes.
   2. Average number of comments.
   3. Average number of views.
   4. Average number of subscribers per content creator.
      1. All of these would be segmented by users who have completed the creator academy vs users who have not completed it (or never accessed it).
      2. Assuming the academy actually helps users get better at creating content, the metrics for their videos should improve.
      3. These metrics would be leading indicators of content creator retention, as creators who are doing well will be more motivated to continue.

1.  According to me, youtube is a very big. So to narrow down the scope I want scope I am dividing youtube into two parts : 
2. Youtube (consumers) : This bascially is the ecosystem consisting of creators and viewers. Creators create their content and viewers watch it 
3. Youtube (business) : This is the business aspect which revolves around the money making side of youtube. Currently youtube makes major money by ads and some by youtube premium. 

1. Content Creators : Content Creators as such makes videos and make money via youtube ads and sponsorships 
2. Business Owners : They use youtube as a source to create brand awareness, to sell products. For that they runs ads or collaborate with content creators.

Currently the process for influencer marketing is very disjointed. Businesses have to interact with agency and they recommend them with influencers for their campaign. This is very unoptimized primarly because : 
 

1. Most of these agency have restriction to demography where as there are many products who have the budget and can be used world wide maybe because they are software or they have huge distribution centre or maybe anything else 
    
2. The way these agency work is very manual. The way they match influencers with businesses is primarly on basis of views and relationships. There is no way for them to figure out what is the sentiment amongst the viewers for a content creator
    
3. Many times businesses partner with influencer that do not resonate with their brand and the pricing isn't standardised it is mainly on the basis of number of subscribers 
    
4. They do not have the techstack to evaluate the success of the campaign.

Solution : 

Building a dedicate platform on youtube for influencer marketing just like we have youtube analytics. 
 

1. Influencer Matching : Business will have to answer a basic set of questions like (Demography they want to target, what there product is about and similar questions) and then our algo will recommend them with set of content creators that fall under their  choice.                             (Reach : Very High , Impact : High , Efforts : Medium)
2.  Creator Insights : This will provide businesses with unique insights about the content creators like (Share of Voice , average engagement on their videos , insights on their comments). These all can be done by running AI models on the comment and doing data analysis on the metrics of the videos 
   (Reach : Medium , Impact : High , Efforts : High)
3. Standardisation in Pricing : These set of influencers will be in grades eg Grade 1 , Grade 2 , Grade 3 and business can choose how many seconds they want their promotion to happen (30 second , 1 minute ....)
   (Reach : Low , Impact : Medium , Efforts : Low)
4. External Linking : Content creators who are a part of this platform can also link there twitter , instagram accounts so that the collaboration isn't limited to just one platform. (Reach : High , Impact : High , Efforts : Medium)

Describe the product & mission of Youtube: 

So Youtube is a video-streaming platform that connects content creators with viewers, and is available across all platforms. 

Clarifying Questions: 

1) What do we mean by improve? retention? engagement? usage? 

2)Are we looking to improve a certain version of the platform? Meaning - web version, mobile verson? 

3)is there a specific timeline for improving this product? 

Interviewer answers clarifying questions - then I let them know based on the informaiton - I am going to take a minute to think about the problem. Before I do, i let the interviewer know, the way i am going to approach this problem is by 1)listing some potential user segements we can focus on 2) select and list user needs of this specific segement 3)select one pain point, and then develop some potential solutions to that specific pain point 4) select one of the ideas and then define some success metrics for the solution. Is this an okay way to go for you, do you have any questions or concerns with this approach? 

Ok so a few user segements that we can look at are users based on types of content being watched:

1)tv/movie viewers

2)podcast viewers

3)at home viewers

4)tutorials 

Before we look at these four, I'd just like to insert that before this step, I first looked more broadly at the user segmeents of youtube - that being 1)viewers 2)content creators - after selecting to focus on viewers - i then segemented the user based on type of content being streamed - which is where we are currently at. 1. tv/movie streamers 2. podcast streamers 3. at home viewers 4. tutorials 

Is there a specific user segment that you like to focus on? (No) 

Okay, then, I would like to focus on the user segment - podcast viewers because this form of content has seen enormous growth of the last ten years - with the increase in # of podcast channels being made. 

Let me take a moment to think about some pain points for podcast viewers. 

Okay - so here is what i am thinking about the user pain points for podcast viewers. 

1) not being able to locate a certain topic within the episode being watched

2) dont like - audio only based podcasts

3) not being able to locate podcasts with a certain guest 

In thinking about these three, I want to select the user paint point that is most frequent to this user. If we can solve an issue that is high in frequency for this user segment, we can potentially achieve our goal of increasing usage of the platform. Would you like me to discuss with you my evaluation of these three pain points - or for time purposes - give you an explantion behind the one i chose only? Okay great - 

1) I think this is medium impact. As a user of youtube myself, it can be frustrated not being able to locate a certain topic within the episode - and happens to me 1/2 times when listening to a podcast. this could be because the podcast host does not timestamp their own podcasts. I would rank this pain point as medium frequency

2) I think this is low impact. Let's say we have a sample size of 100 people, who are podcast viewers, i would assume that half are there strictly for the content itself. Because of this, i would argue that this need is a smaller frequency.

3) I think this is low impact as well. In my experience, while searching for a certain podcast host with youtube, i assume that i am 75% success rate in locating podcasts with a desired guest.

In evaluating all three options, i think the pain point of podcast viewers being unable to select a certain subtopic within the podcast is the most frequent out of the three. 

Now, i am going to take a minute to think about some potential solutions to solve this problem of being unable to locate subtopics within a podcast. 

Solutions: 

1) ai generated timestamp feature for content creators

2) new policy for content creators to publish their timestamps before posting

3) bookmarking feature that lets you organize content based on untimestampped podcasts and timestamped podcasts. 

In evaluating these three ideas - the approach i am thinking that would be most effective is to select based on the criteria of impact and cost of implementing the solution. 

Lets take a look at all three

1) high impact (low cost of implementation)

2)low impact (high cost from a user retention perspective)

3)medium impact (extra work for the user)

In selecting the first, AI generated timestamps, I would now like to lay out some success metrics we can look at to see if this solution implementation will be successful in our goal of increasing users on the platform. 

#1 - # of minutes watched on podcasts per day 

In summary - our initial goal was to increase user engagement on our platform. We selected to focus on the user segment of podcast viewers and implement a solution that addresses their need of timestamps in podcasts. The solution we came up with was a AI-generated tool in the backend, that will timestamp every podcast that goes on youtubes platform. 

Before we try to improve Youtube, I'll first recap my understanding of the business model to make sure we're on the same page.

Business Model Summary
Youtube shows videos to users across multiple platforms, web, TV, and mobile. It is primarily showing youtube specific videos, but there is also Youtube TV which streams live television.  The business model for Youtube is primarily selling ads, but there is also a premium service with less / no ads, as well as Youtube TV which shows live programming at a much higher pricepoint.

To improve Youtube, we first need to know what we're trying to improve.  At a very high level, we have 3 options:
1. Increase User Engagement (# hours per user per day)
2. Increase # Users (# monthly active users)
3. Increase profit (total revenue minus cost)

My inclination here is to focus on #1, as total # hours goes up, ad venue will certainly go up as well.  Is that ok with you?

Interviewer: Sure, sounds good.

User Profiles
Ok, so when we look at trying to improve total # of hours, lets look at some different user groups that watch videos on Youtube.

1. Young kids watching videos.
2. Adults killing time for entertainment - longer form, like from a laptop or smart TV potentially spending a couple of hours
3. Adults killing time for entertainment - short form, from a phone, waiting in line at the grocery store, etc.
3. Adults trying to gain specific knowledge.  Often short form. They will prefer the most efficient way to learn the information, like how to jump start a car.
4. Adults / kids watching live programming on Youtube TV. This is notable in that with this, Youtube is competing with other mediums to watch the same content.

I'm inclined to tackle the young kids demographic.  Young kids have a constant need of attention, so even if the rest of a parent's life is busy, the child will usually not be a thing that gets ignored.  Compared to their own entertainment and youtube watching.    This means that having a good product for child entertainment has a massive opportunity.  Competitors in the market do exist, such as Netflix for kids, but those products are also offshoots from the company's main product line.  The fact that Google has so much information about its users, more so than Netflix, positions google to serve better recommendations over time than Netflix does.

For those reasons, I'd love to proceed on a feature to improve the Youtube for Kids user group.  Does that make sense?

Interviewer: Sure.

Pain Points for Kids as Viewers
Ok, so lets go into some pain points for this user group.  And I think when we're looking at the users being kids, we need to consider the kids parents as well, when looking at pain points:

1. Hard to find the right content. There are so many content options, its hard for parents to find the right content for their kids. If the median video length is, lets assume 8 minutes, if it takes 2 minutes for a parent to find or assess a video, thats a ton of the parent's time. It can be hard to find good content for "set it and forget it" type of programming for kids.
2. The kids as viewers need different levels of autonomy as they get older.  A 3 year old might not be able to choose any of their own videos, but a 12 year old certainly is.  It can be difficult to get the right UI controls for the right age of child.
3. Kids get older.  This means that even if we have an algorithm that intelligently learns a user's preferences, those preferences will without a doubt change over time, because the kid is getting older.  Adults preferences change as their taste changes, but kids preferences change both as their taste changes and as their maturity level changes.

I think all of these issues are related, and get to the core problem that its difficult for the content provider to keep up with the changing needs of the viewer.  Its hard to find the right video, and even if you do, its hard to find the next right video.

Product Ideas

I think we could solve this by building child user profiles, and suggesting playlists based on the user profile information.  We can have either the user, or the parent, fill out a short form as they setup their account that captures age, and interests.  Then we can use this to surface content to the user when they turn on the app.  Knowing the age of the child, we can build a couple of key features:

1. Indexing of all videos for age appropriateness - we can train an AI model on age level approprioteness of the video, then index all of youtube on this to know what age this is correct for.  Then we use this index to only show correct videos to correctly aged users.
2. Show a range of videos based on age of child. If a child is 5, we can show them 50% videos meant for 5 year olds, 25% for 6 year olds, and 25% for 4 year olds.  
3. As we see a childs usage and completion of various videos, we can update our profile for them. If a 5 year old watches all of the videos for 6 year olds, but never completes videos for 5 year olds, we can likely reduce the number of videos for 4 or 5 year olds that we surface to them.  
4. Auto increase the expected age level as the child gets older.  As the child gets older and turns 6, we can gradually update the video suggestions to being 50% for 6 year olds, 25% for 5 year olds, 25% for 7 year olds etc.
5.  Child video consumption dashboard for parents.  As a parent of a child, I might want to be able to see all of the content my child has watched. We can build an admin dashboard for the parent that shows the videos watched, as well as a heatmap of the topics and age levels of the videos being watched.  We can also use this dashboard to set hard restrictions on what the child can watch. We could do this using either an allow list, or a block list style approach.

Prioritization

Building all of the above features would be a heavy engineering lift.  I would prioritize them by weighing impact to the core problem statement against effort to build.  Based on that, I believe feature #1 is key to the entire project, and would need to be done first.  Without that, everything else would fall apart.   After that,  building the onboarding for profiles to get user ages would be the next step. Based on that, we could easily build features 2 through 4, in order. 2, 3, and 4 gradually increase in complexity, so I would start at the beginning then iterate over time.  Feature #5, the parents dashboard is an interesting aside.  The visibility into what their kids are watching might be very important, though the allow list and block list is a significantly different lift from just visibility. Its fundamentally the difference between giving parents "read-only" access to what their kids are doing, vs "read-write" access.  While there is significant value in giving parents the ability to block some content, I think that is less key to the success of growing the kids watching youtube demographic, so I would keep that feature last.

Summary
In trying to improve Youtube, we could do it many ways. We decided to focus on increasing user engagement, and measure that by total # hours watched per user per day.  With many different user segments, we selected kids as being one with a lot of room to grow. Many of the pain points to kids and their families center around finding the right content to watch, so we suggest indexing all youtube videos by age level, building user profiles that intelligently suggest videos for kids based on the kids age, and the age level of video, and tracking changes to the user's preferences over time.

Clarification : Geography : India , Platform : App , Any specific goal ? : Engagement in youtube

Company and product : Part of Google , OTT platform , User create content ,view content .Free ,

Product Goal : platform that helps to upload and view video content.

Business Goal : Improve engagement ( More and users should interact with the product. )

Segments in which youtube operate

Video Content

1. Content creators : upload

2. Content consumers : Consuming content

Content ( time ) :

1. Short form content : less than15 mins

2. Medium form content : 15 mins - 30 mins

3. Long form content : 30 mins+ content .

Content (Genre :)

Entertainment , Shopping content , Gaming , Learning , etc :

Focus on content consumer group.

Users : Who are the users ? 

Demographic Segmentation :

• young adults ( less than 24 years)

  • male

  • Female : Fashion / shopping / education part /

  • Characterisitc : short form content / tech savvy / Learning aspects / enough time as well

• Adults , 24-50'

  • Married

  • Couples / single working professionals

  • Enough money to pay

  • Family with kids

  •  

• Elders 50 plus 

User journey : How and when are they using the product.

open the app

search vs browse

View content ( few recommended content )

Click on content to view

Content viewing experience

Leave or view more content.

Problem areas :

Discoverability : mental effort to find content. ( they might have some content in mind or dont have content in mind)

Effort to search content : physical effort

Which content is the best content to watch

I have limited time to watch - what should i watch

How do i satisfy my knowledge needs / or any aspirational

lots of time :

How do i show off to my friends about the content iam viewing .

Solutions :

• Content rating / of the different videos ( have content in mind) - Discover best content - P1

• Recommendations based on time .

• Show content popular content in your city - New content

• Create a different section of Young Teen section - P1

• Import user google profile to get recommendation of what your friends are watching and liked

• Focus on content which Teen genre like - education / Teen content.

Prioritize based on effort and time taken to complete it . 

Success metrics :

Video time watch time . 

Number of videos watch.

Clarify:

1. Do you have specific goals for this improvement based on any areas such as user engagement, user retention or stickiness for the app or monetization strategy for YT? No
2. Youtube offers several different services such as YT app, YT kids app, YT TV, paid subscription, do you want to focus on any specific area for improvement? - Let's pick YT Kids.

Restating the question:

YouTube kids is a separate on demand video streaming platform that offers safer access for kids on both android and iOS. Liked for it’s simple yet practical design, safety features, parental controls. My personal fav is screen timer.  Already has 35M active weekly users. 

User personas: kids under 5, 5-10, 10+, parents, care providers such as nanny, teachers, extended care staff. 

I am going to focus on the kids under 5 user persona for this exercise because the learning and impression that kids build at this age, shapes and forms their future. From the App usage perspective, I can also see this age group as the one with more screen time. As kids grow older, they have homework, school, extra curricular activities hence < screentime. 

Features:

I can think of two goals for YT kids. One is increasing product usage, adoption as well as monetization.

To increase product usage, there are around 2.2B kids in this world. Let’s assume 25% of that is 0-5 age group. This brings us 550M users. Youtube Kids currently has 25M, active users. So there is headroom for growth.

Features: default app installed on the Chromebooks distributed to school, there are around 2B of kids in developing countries, and identifying a way to distribute a Chromebook even for each classroom in developing countries will blast the user count towards next 1B users, in one way or other.

For the app improvement, I would recommend-

1. Features that enable parents to design a curriculum for the child. Provide different path recommendations based on the child’s interest such as music, space, gardening, swimming. So parents can maintain a focus, increased engagement. Will attract parents/ kids who do not want to watch random videos on screen.

2. e-Commerce is obviously integral part of social media, Youtube Kids can offer functionality where kids can add products to their wishlist based on the ads seen which is available to parents. - monetization

3. Instead of limiting the usage to online book readings, music and other videos, YT can offer live shows as paid subscription, I can totally see parents paying for once a week magic show that offers 

mission: 

anyone can create and share videos

youtube is a mature product. it is in the monetization state. 

Competitors are 

Instagram reels and TikTok in short video space. 

in long video space, there are a lot of OTTs 

Stakeholders

1. entertainment seekers - movie watchers, songs watchers    

2. knowledge seekers

3. news seekers

4. sports

 2. news channels

3. knowledge video makers (brands and independent)

4. independent content makers

1. brands that advertise

improvement areas:

1. gain a new set of users (Gen z) 

2. plan for a new set of entertainment (new content)

3. new ways creators can create something easily 

4. make something for advertisers 

5. new monetization model?

i wanna choose advertisers:

1. because they have pain points in their journey

2. they're being aggressively pampered by the competition (tiktok, meta)

solution:

1. clickable ad with a product in it -> takes to the in-app browser

2. clickable ad with multiple products of the brand -> swipable cards -> takes to the in-app browser

3. let youtube create dynamic ads for the brands of the product instead of brands themselves making the videos. brands upload the products, and let youtube create ad videos.

Metrics (if the hypothesis is successful)

this could improve revenue for youtube through ads or 

increased revenues through youtube premium

counter metrics

this could lead to decreased avg time spent

1. What does the product do 

Youtube is a video services allows users to create an upload videos that are usually 3-15 mins long. The service is free, anyone can upload a video. You can also create youtube channels that users can subscribe to. You like a video, share a video, comment on a video. 

2. What is the goal of this improvement? Do we want to increase revenue, increase number of youtube users or increase user engagement? Let's say we want to increase engagement

3. Who are the users of youtube? There are 2 types of users - one who create and upload content, others who view the content, like ,it share it or leave comments. I will go with users who create content for youtube

4. User journey 

1. User creates a video through some video creation tool 
2. User uploads video in youtube
3. User adds video name, search tags and other information
4. User publishes the video
5. User tracks # of views, comments, likes and shares 
6. User gets paid by youtube if the video hits a certain number of views 

Content creation - this may not allow all types of content creation like lightboard series but should work for a large number of videos, some examples are  

• User can record a video through the youtube application using their computer microphone and video
• User can upload a presentation in the application, and record a voiceover video for the presentation
• Use can upload images,  short videos, etc and mash them up to create a video with voiceover

• User can trim content
• User can add effects
• User can re-record a portion of the content 

• #of users using the content creation app
• # of videos created per month 
• # of videos published per month
• Avg #of videos uploaded per user
• # of active users (viewers or those who have an account but don't use youtube often) who are now publishing content
• Avg length of content created using the content creation and editing tool

The goal of YouTube is two-fold:

1. Give content seekers a platform for finding and watching videos they will want to engage with, learn from, or generally find useful or interesting. 

2. Give content creators a platform for creating videos that a lot of users will be able to watch and engage with. 

Youtube has billions of users, but it will need to increase both revenue and engagement in the future as it competes with other social media and video content creation platforms like TikTok and Instagram. 

1. Content Seekers. Although traditional video content and social media platforms have historically catered to content seekers, i.e. providing a good user experience and giving personalized content to content seekers. We are entering an age of social media where empowering the content creator has led to success, which leads me to…..

2. Businesses. Businesses have used social media platforms to target content seekers for their ad campaigns. This has worked well especially since social media platforms give curated and personalized content to a user base so that businesses can work with users that are either already familiar with their products or more likely to buy their services and products. 

3. Content Creators. The great thing about content creators is that they are both content seekers and in some way businesses. 

With these user groups in mind, in order to improve Youtube here are some metrics we can track to improve Youtube:

1. Content Seekers:

   1. Average time spent watching video content per day -

   2. Average number of likes per week - 

   3. Average number of comments per week - 

 the assumption is that a higher average time spent on watching videos and/or a higher number of likes and comments (aka  engagement)  equates to a better experience for content seekers as they are enjoying watching videos and are finding content that they like

2. Businesses:

   1. Average number of clicks on their ads per day

   2. Average ad conversion on a given video

The assumption is that the higher number of clicks on an ad and ultimately ad conversion (aka Ad Revenue) means that a business is targeting the right videos on Youtube to generate revenue. It is important to note that although ad revenue is great for business, it can disrupt the content seekers’ Youtube experience and there are tradeoffs to consider when determining the number of ads. 

3. Content Creators:

   1. Like Content Seekers, engagement metrics are important as the number of likes/comments/views on content is beneficial for the goals of the content creator

   2. However, unlike content seekers, continued content creation needs to be incentivized and like businesses revenue metrics should also be tracked. 

Since I would like to target content creators for this discussion, I would like to choose to focus on content creators (as this has led to the success of other platforms like Tiktok) and focus on engagement metrics since engagement can lead to fame, recognition, and possibly monetization of content. 

The main problems Content Creators may face today on Youtube include:

1. Lack of advanced video editing features in Youtube. For example, video editing features like filters, music editing, photo or video layering would make it easier for content creators to create on Youtube. Other advanced features like video templates and text, music suggestions are lacking today.

2. Lack of content creation analytics or dashboards such as knowing which time to post and hashtags to use to target specific audiences means content creators may have little awareness of how to target specific content seekers and when for maximum engagement.

3. Lack of methods to monetize engagement for content creators. Monetization features can incentivize max effectiveness content and continued creation of content. 

Some potential features to address include:

For an MVP improvement of Youtube. I would first start with providing monetary incentives to Content Creators and a “Payment Dashboard” that content creators can use to track their views, how much money they received, and suggestions to create more monetary optimal content for their targeted content seekers. This would provide the highest impact on engagement as money Content Creators received would be directly tied to engagement (views/likes/comments) on their content. In iterative phases, I would make it EASIER for Content Creators to make content by adding more Advanced Video Editing features and lastly would focus efforts on an Analytics Dashboard that gives Content Creators data points for improving engagement on their content.  

Implementation of MVP would require the following considerations:

• A secure account for peace of mind and fast transactions between Youtube and Content Creators

• A way to make financial transactions work wherever a Content Creator is located

• Possible API integrations with PayPal, etc.

For Advanced Video Editing and Analytics Dashboarding would require:

• Building accurate algorithms that make use of Youtube’s large database to make editing and content creation suggestions.

• UX design that makes it easy to make edits and deploy content quickly

Improve You tube

Describe the product

Youtube helps in sharing the content in form of recorded video in a user friendly and interactive manner. It is mostly used by the individual users like students, learners, entertainment purpose, knowledge and the content creater who wants to gain revenue and achieve publicity amongst the masses. Assuming, we would like to improve youtube in every platform be it cell phone, web, desktop.

Goal of the product

The goal of the product is to increase the customer engagement, acquire new user base, improve retention rate. Assuming we would like to improve customer engagement.

User Groups

User group A: Business users- who creates the content for publicity and generating revenue

User group B: Individuals- Who wants to see the content for education purpose, know more about the new things, entertainment

Age group for user group B

User group 1: Kids and youth 5 years to 35 years (Star users - who uses the you tube very frequently almost one in every hour)

User group 2: Middle age- 35 years to 50 years (Active users- who uses you tube as per need neither too much nor too less, may be once or twice in a month)

User group 3: Old age - 50+ (Dormant users- who uses you tube occasionally, rarely)

Pain Points and use cases

1.      User group 3- Inability to find the desired content quickly leads to distraction

2.      User group 1&2- Protecting the kids from the age-inappropriate content that appears inadvertently

3.      Inability to identify the authentic content based on which some decisions can be taken like stock market updates

4.      Too much fake content. Genuine content has been overtaken by fake/false content

5.      Regulatory perspective- Inability to identify the trail of risky and too much harmful content that can have negative repercussions on the society at a large scale

Solutions

1.      Search optimisation techniques like suggestions as per the user’s past search behaviour

2.      Use of algorithm search like + and or = to narrow the search criteria and get the most relevant search results

3.       A short message or a flag on the offensive and age-inappropriate content to warn the user before it appears. A notification mail can be sent to the secondary ID of parents to warn them if the primary user who is a kid accessing the in appropriate content with detail like time, video link IP address and vice versa

4.      An alert mechanism to warn the users on the fake video if it has been manipulated, decorated using some tools and apps. They can still watch it but aware of the fact that the video is not encompassing all the parameters to meet the authenticity. Like 3/10 a rating system

5.      Tracking the location, IP address to the support team if too much harmful content is released so that preventive and corrective action can be taken as soon as possible and rewarding the content creators who shares authentic excellent quality meaningful content apart from regular revenue generation. This could involve reward points and badgyfying the authentic good quality content creators

6.      Integrate the youtube specifically to the site like linked in etc to get the benefit of large network and allow them to share the required content

Evaluation

Solution 1,2- is high in business value medium in efforts and cost so it should P1

Solution 6 is high in business value high in efforts high in cost but it will help in utilizing the available network. The pitfall will be if the deal is denied then it will not be possible to embedded this feature

Solution 5 is medium in business value, medium in efforts and medium in cost as it is not that difficult to build but it is critical for even one single corner case

Solution 4- is high in business value, lower in effort in cost and it is very much needed to gain trust from the viewers

Product description - Youtube offers video streaming solutions to all age groups. UI primarily in English. Video content creators can monetize the video(views, likes, comments, subscriptions). Revenue - Ads.

Does this sound like the right description?

Clarifying questions -

Are we thinking of any specific objectives in terms of improvement?

(Assuming it as an open-ended question)

Solution-

Scope restriction:- Youtube is an enormous product with a large user database. It also has a user database in multiple countries(almost all across the globe). We can improve such product on several means or ways.

I am trying to target particularly the User retention aspect for this exercise.

Q - Is that okay or shall we take another element?

Assuming that interviewer gives a nod to proceed.

Before jumping to the solution, I would like to understand the user groups. like any other product of this size and complexity, youtube also has power users, regular users, occasional users.

Having said that, user retention use cases can imply to all user groups. Just for reference, I will try to analyze the user story of a power user(access youtube multiple times a day or spends a minimum of 1-2 hours on a daily basis).

Pain points/Improvement area identification:- 

1) videos are scrolled. user has to click the full-screen button most of the time.

2)Once the video is finished, the user has to click - back button or close the full-screen button, or sometimes even close the minimized screen window.

3)Youtube does not allow to create content instantly unlike other competitors do.

4)Users moving to other reels/stories applications that are more convenient to operate.

5)Forced ads causing users to slowly dislike youtube and eventually leaving the application.

6)Content deficit of a particular interested area of a particular user.

7)Users leaving the application as they realize that they are getting addicted to the application.

Let's try to prioritize by analyzing a bit more.

1) UI-related - videos are scrolled. user has to click the full-screen button most of the time. Once the video is finished, the user has to click - back button or close the full-screen button, or sometimes even close the minimized screen window. - Medium Priority - As youtube primarily is a video streaming (regardless of video duration), but not a short video streaming application, there are some differences in UI. The youtube shorts module does not have these problems. Also, we don't want to become a full-on short video streaming app. 

2)Youtube does not allow to create content instantly unlike other competitors do. - Medium to low priority - Not all users are content creators. Thus, the target user base is less.

3)Users moving to other reels/stories applications that are more convenient to operate. - losing customers to short reels/stories application - High priority- since new applications are coming every day for short video streaming, increasing trend of short videos.

4)Forced ads causing users to slowly dislike youtube and eventually leaving the application. - Medium priority 

5)Content deficit of a particular interested area of a particular user. - Low priority - application usually has a huge database for any area of interest.

6)Users leaving the application as they realize that they are getting addicted to the application. - Medium priority - A section of power users usually can get impacted due to this pain point. The Target user base is relatively less. However, this can be a potential future pain point.

I would personally prefer to explore point 'E'. 

However, point 'a' is something which can be built in quick time and can achieve an almost similar impact. I would prefer to go for 'a' and make it a base(integration module) to start working on point 'e'.

Metrics -

a. For creators - # of videos using new feature / total videos created

b. For users - User engagement on videos with new feature / avg user engagement (Likes, Shares, Views, Time spent, Adding to Liked Videos/Collections) of all videos.

Summary - to improve youtube, I tried to narrow down the scope to user retention and started identifying improvement areas for user loss. we prioritized the pain point that YT short video section is losing users to competitors which are dedicated to short video streaming. After looking for multiple solutions in terms of impact vs effort - 'using google photos/ videos to auto-create the short videos for users'  and 'Integration of google lens/maps' looked feasible and impactful. Finally, we defined the relevant success KPI.

P.S - I would really appreciate the feedback you might have. please let me know what you think.

First I need to understand that, what type/category of user problems we need to solve to improve youtube.

User problems can be classified into different categories like..

 1) Improvement in the Customer engagement metric, ex. when/where a user is dropping while watching a video, some other KPIs  like how much time a user is spending or how many videos a user is viewing every day

 2) UI/UX improvement

 3) Improvement in the search time- Time taken by Youtube for suggesting available videos to the viewer

 4) Improvement in the recommendation engine of youtube, so more relevant videos will be showcase by youtube to a user.

 I am making the assumption here that Category -1 is the scope as of now, which is improvement in User Engagement & taking the use case of Customer dropping while watching the video.

Now after data analysis & interviews with a set of users (low/medium & High you tube users), we created the problem statement that due to in between video advertisements, users lost interest & drop out.

Now while framing the solution, we analyze what is the Current solution to this problem, so the current solution is subscribing to youtube Premium for ad-free viewing of the videos - but for the Indian Market, the youtube premium subscription charges are quite high  (Rs 139/month for 1 screen).

 The solution I am proposing is the introduction of the Loyalty program, where a user will earn points by watching videos and they can redeem these points by subscribing to youtube premium for a limited period.

Earning Points will depend on

1) Number of videos watched per day -- Ex. 1 point for every video

2) How many videos is completed by the user - Ex. Bonus 1 point for every video completion

basis the past behavior of the user (on a number of videos watched/videos completed) youtube will give the segmented reward of ad-free video.

This will improve the following  KPI 

1) User engagement -Time spent on youtube/no. of video watched 

2) Adoption metric - Post-Ad-free experience, XX customer will upgrade to youtube Premium

Please suggest how you like my approach & please give your feedback. 

To be on the same page, would like to clarify the understanding of the product. As per my understanding, Youtube is the a product where users can create videos as well as search/view/like/share videos posted by other customers. Its like a search engine for videos.

Next would like to understand the objective of improvement, as per my understanding the objective is to increase the engagement or time spent by customers on Youtube.

Next I would like to segregate broad user segments:-

1. Users who are creating content/video
2. Production houses/brands who create videos for promotions
3. Users who are searching & viewing the videos

1. Often while searching for content, Youtube shows the same content which u would have already seen. Thus often u end of clicking on same video which u have seen before. OR customer himself has to think whether he/she has seen the video from the title.
2. On the home page, recommendations of new videos as per customers past watched history are not visible in the first 2 folds on mobile. There are ads, then promo for the new 'shorts' feature, news and then recommendations. Also the recommendations are not aligned as per customer's past watched videos. Thus leading to customer drop offs.
3. Also reviewing already watched videos is also bit cumbersome as u can save videos but not in segregated folders... thus after some time the list of saved videos is itself quite long and searching within it is difficult.
4. While viewing the videos, often customers want to scroll 5-10 secs back or forward.. but this can only be done by manually dragging the ribbon which often is not accurate and quite difficult via mobile devices.