Analyzing YouTube's View Duration Drop | What Went Wrong? | Adobe & Flipkart Experts | TPF

[@vipulmangla2162]

Comparison of KZbin Premium vs KZbin Free would have been an interesting metric to look at when we were looking at Skip Ad button issue.

[@SuhasMotwani07]

Great point!

[@abhishekshuklain]

Spot on! That would have helped us get to a conclusion much sooner.

[@mohammadabdullahansari6314]

@@abhishekshuklain here Anushka could have used 5W1H much better particularly Who

[@SuperFatafati]

Irrelevant, the case statement said that drop was observed across all products. KZbin premium and youtube are pretty similar products, the case statement would have meant issue was faced irresepective of which platform it was on

[@rahuljain4646]

Well explained . Anushka did a great job specifically with the first Q on listing all options. Thanks to Abhishek as well for explaining what else could have been considered.

[@abhishekshuklain]

Glad that it helps, Rahul!

[@sakshamvikram7263]

Few points here External:- There is a hate campaign running against KZbin for any of its recent policy changes. INternal:- What if there is an issue with the tool reporting the metric. Another one can be is search working properly? Any change in algorithm leading to bad results. How many search hits are there?

[@abhishekshuklain]

All great points. The idea is to follow a structure, ask clarifying questions, look at metrics and adjust your next steps based on the response of the interviewer for the metrics. Also, as a tip - try exhausting internal factors first before moving to external ones.

[@bollamani]

One of the best resource for interview preparation. Not only aspirants, experienced PM also can learn some good insights and techniques from this video. Thanks to both Anushka and Abhishek

[@abhishekshuklain]

Hi Mani, thanks for your appreciation! Glad that this resonates.

[@TheProductFolks]

Well said. Stay tuned for the next in the series.

[@KshitijAnand-m1j]

At around 3:05, the interviewer mentions that "one fine day you open the dashboard and see a 40% decline in the metric". Maybe because he stated that, the interviewee skipped the "timing/occurrence" question. I did the same while practising. Otherwise, good stuff.

[@tabishakhlaque4860]

A follow up question here: Button click issue is a major bug. At the same time interviewer clarified that no bug has been reported recently. Contradictory things.

[@nitinshekhar1010]

In the guesstimate problem, i suppose we forgot to account for people travelling and staying in bangalore for 2-5 days.. though not a very huge number, but if we account for bangalore rain and traffic, this could have been considered.. but a very nice analysis nevertheless.. thanks for that

[@abhishekshuklain]

@Nitin, good point. While it’s a good idea to mention all the possible segments/scenarios, it’s completely OK to consider only the significant ones for the guesstimation. You may mention to the interviewer the criteria that you are using to consider or ignore scenarios.

[@tuktuk6740]

this one is really good. She did her job perfectly

[@ashishdimri5660]

Really liked this interview. One question I would like to ask after watching this interview is: What books or resources do you recommend to prepare for these type of guestimates?

[@abhishekshuklain]

Glad that the video helps! "Cracking the PM Interview" is a great book to refer to for a general approach to problem-solving. The best way to get better at estimation questions is to practice a lot of them. Stick to the basic approach - Ask clarifying questions; State and validate your assumptions; Break down the question into logical steps; Solve/estimate for these individual steps and keep validating with the interviewer; Summarize your final solution and any trade-offs. Also, don't worry about accuracy. Focus more on the structure and approach.

[@ashishdimri5660]

@@abhishekshuklain Thanks a lot. For someone who's just starting to learn about the PM world, this advice is very much appreciated.

[@nsubhakartubes]

@@abhishekshuklain Any suggestions on resources that can help on in practicing estimation questions?

[@abhishekshuklain]

@@nsubhakartubes You can find some questions to practice here - www.productmanagementexercises.com/interview-questions/estimation

[@nsubhakartubes]

@@abhishekshuklain Thanks a lot!

[@yashanilsaboo5662]

I think the number quoted is high because you've considered the frequency based on months but the question as I remember was for weekly order number, soo the final answer could be divided by 4 or weekly orders

[@AdityaMba1988]

Interesting, informative and great insights. Thanks Abhishek and Anushka

[@Sachin.Rastogi]

We should also consider the popular social media sites, e.g., Reddit and Twitter, for understanding the issues users observe across the Globe.

[@nsubhakartubes]

Such a valuable video. Thanks to the interviewer and interviewee.

[@abhishekshuklain]

Hi Nithin, glad that you found it helpful!

[@TheProductFolks]

Glad you enjoyed it!

[@EklavyaMishraJordan]

Good learning to take away after watching.

[@arpittrivedi8458]

I think - the lack of personalized newsfeeds based on the users interest could also be one of the factors which leads to the huge decline.

[@rahulmaurya5994]

@@arpittrivedi8458 that would reflect in no of users starting a video. As the interviewer mentioned there's a spike in this metric so the above hypothesis is unlikely to be true

[@ruchapallewar5843]

Loved Anushka's approach!

[@meghshah37]

Few questions in the guesstimate section that would change the numbers quite a bit: For working professional many of them would have hired cook. Also many would be eating in the office premises as most of the startups/IT companies offer food. So in my view the order per week for this group of people should be 3. Does this assumption holds true or I'm over engineering/thinking the solution?

[@abhishekshuklain]

Hi Megh, that's an interesting aspect and an important factor to consider if you believe that the segment (working professionals with hired cooks) is significantly large. For most T1 and some T2 cities in India, this would be a significant population and an important factor to key into your calculations.

[@anju3692]

@@abhishekshuklain Yeah,even I thought of the same , putting a 3 there first Anushka made the guess. Again two angles come here : before covid and post covid. That can make a change in numbers.

[@sumitkumarsingh1053]

Sir ...may I combine both english and hindi language to answer PM questions or to answer question in english is compulsory??

[@abhishekshuklain]

Hi Sumit. It’s advisable to stick with English unless: 1. The interviewers themselves use Hindi intermittently while conversing with you. 2. You share a rapport with the interviewers, possibly prior to your interview, and you are confident that they are comfortable conversing in Hindi. In any case, if you wish to use any language other than English, it’s always better to ask and seek agreement.

[@Instantalignment]

Good guestimation! By now she must have become PM.

[@saurabhkhaneja3366]

one other metric which would have reduced the avg view size is the video quality and is video getting stuck. If users are trying to open a video and it keeps on loading, then either they wait or try to restart (increasing the no of clicks on video). This could point out to streaming servers issues

[@abhishekshuklain]

Yes, that's a possibility. It's important to validate this with the information/data you have. For instance, it's unlikely that this infrastructure (streaming server) issue would cause a global impact. Over the course of solving the question, you must eliminate possibilities based on your validated assumptions.