You're the PM for Facebook Like. You notice that Like acceptance is down 5%. What do you do?

PS: I hve no clue how this is going to turn out considering this is supposed to be a very conversational case and I am going to have to really play a true dual role.

Me: Just a few clarifying questions before I diagnose the problem if there is one. After I am done with some cursory look, I will let you know what approach I mgiht want to take. Like has 2 parts, 1 is the plain old Like that FB had for years and then there is also the emoticons (reactions) part. Here, when you say Like do you mean Like/Reactions or just Likes? (all of it).

Ok, so when you say Like acceptance, you mean people clicking or responding via like. (correct).

Is this down worldwide or just US? (just US)

That's definitely a key data point I feel. I would want to keep that in mind here. Some questions I ask may orient around this data point. But for now, let me understand this further. Is this down suddenly or over a course of period? (just since about a week).

ok so that's another critical point to keep in mind. Right off the bat, what I am thinking is there could be some seasonality here. So, is this noticed just this year or we see similar pattern every year? (just this year).

Alright, so no seasonality here it seems. I think there are some other important pieces like OS, device type, how we measure, etc. that i want to cover first before diving deep just to understand some high levels about this. (sure).

Is there a specific device where we notice we don't get likes from? Scuh as people not using like on desktop/laptop vs. mobile device or we see a drop from iOS vs. Windows? (no). Alright, in terms of measurement, typically there is DW activity in the backend to obtain data from OLTP systems, transform data to OLAP and report. Are we sure we haven't changed our reporting system, way we calculate or maybe heck we just mis read the data? (all of these are not correct in our case. Problem is hence somewhere else).

ok...I think this gains me some cursory understanding. So what I would like to now do is explore some user journey / funnel aspects to see if the problem might be somewhere else, then think of various areas such as: code changes, new feature implementations, maybe some environmental issues, competitive actions, reddit forum type of hype / news circling that could affect user action, any other news that could be impacting us, etc. Maynot cover all aspects since I don't know what I am going to learn as we explore this. Are we aligned? (yes).

While the likes might be down it's important to understand if other engagement is happening or not (such as people may just not be logging on or opening FB to evne like so then we would see similar drops in Comments, visits, sessions of FB, etc.). Do we have similar drops elsewhere on FB or is it centralized to Likes? (it's just Likes).

ok, now that we have checked the entry point of user journey, do we see less scrolls through various news feed? (no). Ok, related to that, it could be that maybe we just have less # of posts from creators. Actually, leads me to ask a cursory level question, do we know if the Likes is impacted for certain category of posts or just any and all (posted by business, by friends, etc.)? (it's for any and all). ok so back to my question: do we generally have less # of posts now? If that's the case then it's not likes it's just that something else is driving less posts so people just don't as much material to like. (no # of posts haven't been affected).

ok, having checked user funnel, I want to think of other aspects. You had mentioned that we notice this just in US. (correct). So before I explore some environmental, infrastructure sort of things, I want to first ensure if there were some changes we made because that could impact. (what type of chagnes are you thinking about?). Good q. The type of changes might be our AI algorithm resulting in less interesting material being presented to user, UI enhancements, maybe limitation on # likes (I could see a reason why this could be done), changes to other parts that could have introduced an unexpected bug resulting in perople not able to click like, etc.). (no such changes are at play). Great! Also, cursory level clarifying question...is the like impacted just from notifications or overall. (overall).

If not code changes, do we know any particular part of the nation we have noticed this drop? (no). So that eliminates my question about maybe a natural calamity like Katrina or infrastrucutre related like VZ has had a major network outage or Austin just had a major winter storm, etc. (correct). Ok, did we have a new competitive entry that is driving users away? I assume not since you didn't mention our engagement has gone down. (correct). Alright, do we have an uptick in mention of FB in other products such as Reddit for instance? (don't know). Can we please pull up data? (sure).

Interviewer: Looking at the data, it seems FB is mentioned a lot more than expected. It seems we have been mentioned 25% more times in this last week then ever before over course of a year.

Great! so what type of mentions do we have more? (It seems that Likes and Reactions are mentioned).

Great! what I would like to say is that it's likely there is negative sentiment around these mentions on Reddit. If we have data on sentiment we can dig into it right now but since we don't I would like to work with data scientists or yourself to dig into it. Negative sentiment around it could casue people to abruptly stop using but not all will stop which could explain why 25% of increase in mentions but only 5% drop.

Interviewer: very well. You can book a time with me on my calendar for next week and I can present you some data analysis.

Thank you. If you are fine, I would like to park it here. Problem could be somewhere else and that could also be mention of bad repo in news. Do we have any of those? (No, I think this might be our culprit. Thank you for your time).