How to Answer: Execution and Metric Tradeoffs Lesson (Product Manager Interview Question and Answer)

[@LiftoffPM]

👋 Try 1 paid lesson or unlock the full course at: tinyurl.com/LiftoffPMCourse Hi there! Need help acing your PM interview? Our LiftoffPM course has helped 90+ PMs land roles at FAANG companies and top startups (from APM to VP Product). Click the "Join" button to unlock our frameworks, strategies, and approaches to ace the different types of questions you'll face from execution to product sense to strategy to behavioral/leadership! 📝FAQ: tinyurl.com/LiftoffPMFAQ

[@pankaj_singh_negi]

Thanks! 👍🏼 Please post more often. Very helpful.

[@bbss718]

in the music streaming app example, the metrics tradeoff question and the root cause analysis example you stated are exactly the same.

[@pakouvang-atkinson349]

I've watched so many videos, read through a lot of resources, met with coaches, and you have given the BEST explanation of tradeoff questions. I love the details on the variations. You even gave a clear example of the key differences between RCA and tradeoff that can sound similar and beautifully explained why it would be one vs. the other. Well done!

[@LiftoffPM]

So glad that you’re enjoying our lessons - we have more to come!

[@9094daniel]

Hi Kevin, your videos are on point and I hope I get to your point one day. Thank you

[@rickhouten1622]

I feel this is one of the better videos on specific examples of Tradeoff subjects. Just go into how you should tackle it and called out the nuance between this and a very similar root cause analysis question

[@chinu1111]

Hey Kevin - great insightful video ! Just one confusion I was hoping you could clarify: 1. For the music streaming app example, @8:18 "Discuss decision making process", what exactly is Treatment and what is Control that was defined in the original experiment? 2. Is Treatment = Internal Shares launched new, Control = External Shares ( existing ) , or is it some AB was launched for the music streaming app, and [ internal shares up, external shares down] are the results we're seeing in Treatment ( whatever that is for that feature/AB )

[@akj7799]

Excellent Structure. Is there a reason to pick revenue based NSM for all three examples? Would love to see you try something more complicated for us to understand how you navigate such waters.

[@ghazanfaralikhan4415]

love the content. Audio is low compared to other videos on KZbin and I can barely hear at full volume on my mac

[@LiftoffPM]

Thanks for the feedback - will try a better mic for future videos!

[@janekim279]

Question regarding Target: given revenue and LTV is a lagging indicator, would it not help to define the north star metric as # of items sold per week as a leading indicator? And then design the experiments that can push that? What’s the rationale behind keeping revenue as the north star metric (besides the fact that the interviewer is providing that context)

[@LiftoffPM]

That’s a great question! In this case like you mentioned, the interviewer gave that context (real question we’ve gotten asked), but in another version of this question - where the interviewer didn’t set the target already - # items sold per week would be very reasonable. Like you said, leading indicators are typically better if the interviewer gives you that space. PS: We are planning on making a video specifically for defining NSMs, too!

[@Rico1ii]

hi Kevin, great video! I'm a little confused about the TikTok example. There seems no connection between the hypothesis and the treatment.

[@LiftoffPM]

In that example, we posited that a specific cohort might not value real-life connection on TikTok since their real-life connections exist on other platforms. To test this, we removed/swapped out follower suggestions for sponsored posts and to see how these changes impacted the metric we’re optimizing for. If the hypothesis is true then we would see an increase in our metric.

[@Rico1ii]

@@LiftoffPM Thanks for replying, Kevin. I might have focused too much on the age differences mentioned in the hypothesis, which led me to think the experiment was targeted at a specific generation, while it wasn't. Now I understand. By the end, we can look at the metric changes across different age groups, verify the hypothesis regarding millennials, and possibly take the next step to optimize further based on that.

[@LiftoffPM]

You got it :) thanks for clarifying with us, now future viewers can see this!

[@sahil_190]

This is good I didn't understand the connection at @7:42, are you reverse engineering the A/B test from the data in question provided?

[@LiftoffPM]

The goal here is to look at the north star you've defined, and think of the metrics that need to go "right" in order for you to drive that north star metric. So yes, in a way you're reverse engineering the list of what submetrics need to succeed in order for the north star to grow.

[@LipikaMukherjee-l2b]

Hi Kevin, great video. I had a small question though. Since the app is in the early growth stage, wouldn't the focus be acquisition rather than monetization?

[@LiftoffPM]

Which app are you referring to?

[@LipikaMukherjee-l2b]

Hi Kevin, great video. I had a doubt: Since the app is in the early growth stage, shouldn't the focus be on acquiring new users than monetization and having paid subscribers?

[@LiftoffPM]

Hey, thank you! Which example are you referring to?