The recruiter email says that during the onsite interview a member of eng team will “evaluate technical competence” and to be prepared for a whiteboard coding question. Honestly this surprised me. Do I really need to practice leetcode for a Product Management interview? I absolutely agree that PMs should be technical enough to influence and have credibility with the engineering team, but I would have expected the questions to assess that in a different way than whiteboard algorithmic questions. Thoughts? Should I really be practicing some leetcode or can I expect other types of questions assessing technical competence (eg describing architecture of past project, system design question, ...) 6yoe at MS
I have had two onsite PM interviews. One had no pseudocoding, the other did. The one that did was literally the entire session. I was unprepared for it, and it's likely what tanked me.
If it makes any difference, I had 12 yoe of PM experience at startups. I realized from that interview that I was better off going back into the field ;)
There are three ladders of PM at Google: T - Technical (eng) O - Operations (typical) E - "Business " Product Managers, who were acquihired and don't need a CS or eng bar. Typically, O ladder PMs are expected to have the same technical competency as SWEs one level below them. So an O5 should have T4 software skills. I don't know of many T ladder PMs
Do they ask leetcode hard for PM? Or is it safe to just practice mediums?
Product managers at Google can be very technical, they do write serious tech design docs and also code at times. Not all of them but many can.
I do think I can get pretty technical and consistently get positive feedback in that area from my engineers at microsoft. I also code side projects for fun but that’s different than being sharp on algorithms / leetcode. Would you expect easy/medium/hard leetcode for PM on-site? On a related note I’m surprised none of the example questions included writing SQL queries. Seems more relevant than algorithmic questions to me? Might be totally wrong.