Netflix Product Manager Interview: Inactive Users

Don't leave your product management career to chance. Sign up for Exponent's PM interview course today: bit.ly/3JVhkRr

He's directly looking into our souls lmao

1. Whether this inactivity has been a trend or a sudden drop? 2. Was it web vs mobile specific? 3. Was it android vs ios specific? 4. Was it category-specific - TV series vs movies? 5. Any demographic specific? 6. User journey- from login/signup to discovery to actually clicking the content? Is there a drop in those parts of the funnel? Where are all these questions? People getting bored doesn’t cause a sudden drop. And if it’s a trend, the period of the trend is important to figure out.

Dude does not blink.

Thanks for these videos Kevin/Exponent! There are a few key points missed in the above solution : 1. Is the 10% of inactive users in a particular age group? - This question could lead to insights to lack of content aimed towards that user group 2. Is the inactivity / drop over a week? Or a month? - there is no time frame considered 3. As some folks already mentioned below, "past" watch behavior of the users could point to a lot of useful insights.

My takeaway from this mock interview (Case of Netflix inactive users) - Define the problem. What are inactive users? - Ask questions to build hypothesis/identify the why behind the problem. In this case, try to get answers to questions like - Have the users watched all the shows on My List?, Is there any negative news about Netflix that's kind of floating in the market?, Is there any new content that is published or we have old shows only?, Is this happening for any specific region or globally?, Is there any new product that's kind of giving competition to Netflix? - if yes, what's unique in there? - Is it their pricing strategy/discounts or new shows being published going viral that kind of drifted away our users attention? - Based on the insights we got by asking clarifying questions, come up with solutions. - Decide what metrics you'll track to gauge improvement. Now metrics here can be - 1. North Star Metric - How many users are logging into the app?, 2. Engagement as in are they taking up the quiz or kind of reading the show description, sharing shows with friends, checking out what's new - adding it to My List, etc. 3. Acquisition 4. Retention Note - Throughout the process, do keep checking if the solution proposed is aligned with the problem we are solving.

One question I would have loved to hear was whether 10% inactive users is higher or lower? What is the seasonal pattern. During summer, you will definitely see a drop in usage due to vacations. Second thing is what happens to inactive users, what percent cancel after certain time or what percent come back at different time of the year.

the way I would approach this are: 1) Define the inactive user: I would define active user by # of minutes watched. I don’t be log in a good indication of engagement. For example, me logging in and find all the shows no relevant, and not watch anything is probably bad. Since the value of Netflix are the content, and inactive user would limit to cancellation. 2) Link inactive to business problem: users may not watch Netflix daily, so it’s important define this further. Perhaps, inactive users are those who watch less than 60 minutes in 5 days. Additionally, I would tie this to the business problem. Because if 95% of user who watch 60 minutes in 5 days will keep their subscription may be acceptable activity. I would leverage data to figure this out. 3) Mutually exclusive collectively exhausting categorization of possible inactivity, things like stale content, possible UX changes that drove it (unlikely), possible recommendation algorithm changes, competitive actions that draw eye-balls, possible macro changes, etc. See if there are any clustering of inactive, age group, geography, etc. Then see what the cost benefit trade-off is in trying to change move the needle from “inactive” to “active”.

Would be intuitive to break down the problem at hadnd here, especially drill down on the user cohorts for the inactive users. Could braodly be divided into below categories : 1. New inactive users ( D7 logins which havent been activated yet- not watched any videos ) - could hint at an issue with onboarding ; lack of default recommendations basis region 2. Old in active users ( previously engaged but recently inactive ) - Coud hint at : A. inferior recommendations - check increase in homepage bounce rates ; soln - try push notifications to re-engage basis genre/director/actor prefernces - Note B decrease in subscription renewals - low intent to renew subsctoption-soln - can make case for offers for high arpu watcher who are now inactive/churned

Interesting approach. I think the interviewee could have asked if there were any issues users have reported having with the service or any major trends from user satisfaction surveys.

May I know what was the solution to the problem? I couldn't understand bcoz interviewer hypnotised me.

Hey, great session. Just one question, the solutions mainly talks about how users can login and then check the wraps/new features, but isn't that something which comes after user logs in. The possibility of user logging and not watch a video is very less if he/she is bored on the platform already. IMO, before any of these features, what helps is probably providing them a notification message saying that they watched a certain movie at certain time (that nostalgia bit), to bring the user on the platform (which is the first step in activity) and then probably having further engagement.

Diagnosis of Behavioural aspects of inactive users are very important. Diving deep into past data will give good insight on what they actually enjoyed or watched, Average time they spent on Netflix ( Past) and how many times they logged in but didn't watched anything. Are they get overwhelmed with choices or they get bored with existing content?

Loved Kevin's follow up questions!

Great interview @Exponent and Chloe! I would like to add few more questions: 1. Is this 10% inactive users mostly came from smartphone or tv or laptop users? 2. The 10% come from which type of shows? Cinema, TV series, docuseries, standup, etc?

Great feedback at the end. I am sure at the end, by "activation", Kevin actually meant "re-activation"

I understood the question as the users are already Inactive and it wasnt a drop. So from that perspective it makes sense to analyze these user personas and then figure out solutions to convert the inactive users to active users. Also its crucial to define what's an acceptable inactive user percentage, maybe lets say below 5% is acceptable. Then the solution should be bringing down inactive user percentage to 5%.

Thanks for the video! Few missed points according to me: - Defining inactive and active customer initially - Defining age group, watch time, device used

I didn't understand from the problem whether there was any increase in inactive users or is it a segment size that we want to work on. I would clarify that first.

Is there any model she followed to ask these questions