Facebook Data Science, Analytics IC5 Offer (vs Twitter/Lyft)

Also Data Scientist, at their equivalent senior levels

Do you have offers from Twitter or Lyft? FB one seems standard for the level

both lyft and twitter hire into role/team (at least currently), so the interviews were more thorough and went deeper into stats/ml than the fb analytics interview. - model performance metrics (precision/recall, real world implications of fp/fn with case studies) - classical vs bayesian approaches in experimentation - providing examples of projects you worked on - heavy emphasis on causal inference, probability, and metric selection/validation at twitter, but that's likely dependent on the role/team you're interviewing for. all three companies framed their ds teams as an advisor team that partners closely with product/eng, and all companies seemed to prioritize communication, results, impact, metric definition, collaboration especially for leveling phase. fb analytics role was general pipeline so very high-level and focused on impact, less so on the technical. fb technical had some questions on a/b tests, rollout plans, and two-sided market tests. my understanding is also that a lot of the internal tooling at fb for ds analytics abstracts a lot of the statistics/math away and you're working through guis to run analyses. maybe that's why pay is lower at fb as well

i was also leaning towards twitter over fb since tc is ~100k higher, but last week's hack is making me nervous. surprisingly, twtr stock didn't drop much, but not great to hear while considering a company. good luck if you end up interviewing!

what about Twitter ?

Less than a week

Within how many days, did Twitter revert back with offer after onsite ?

thanks, i was pleasantly surprised too! might be because ds shares eng ladder at twitter do you happen to know how common ic6 is within ds? ive heard there are very few ic6 in ds, and that there's a lower ceiling for growth on the ds analytics team compared to eng and pm. seems like most ic5s at fb ds end up moving to mgr track for growth. id be ok with that, but im curious if the rumors are true that its hard to move beyond ic5 at fb ds

I personally know at least one ic6 ds who was promoted there, and ic6 can probably be earned within 1-2 years. I know of a few m1 hires but no ic6 external hires so far while ic5 is terminal level for many, fb might be good in the sense that most DSes are just product analysts in disguise as there’s a separate research scientist ladder for core research and hard quant. this might make it easier for more technically advanced quantitative people on the DS ladder to stand out in FB but it might depend on the team and a bunch of other factors FB tends to promote early rather than late, so there shouldn’t be much of a worry about getting stuck for reasons out of your control

I can’t tell if you’re serious or not

The TC is guaranteed. There will be no more layoffs.

op here... blind locked me out of my account and i had to create a new one, sigh. i added pointers on what stats concepts to study on an earlier reply, but things that helped me prep and get good offers from twitter/fb/lyft are: 1) drive the conversation and communicate well. you directing the convo signals seniority. make sure you talk in depth about pros and cons of your answers and use real life examples from work you've done. prepare detailed answers to common interview questions and practice saying them out loud. 2) for every job you've had, demonstrate impact and knowledge (helps if it's a long-term project so you can speak to in detail about your process and business results). for ds - get relevant experience and/or keep up with ds industry best practices in tech. data science twitter is great for keeping up with trends. use that to keep learning and make sure you apply some of those concepts in your day to day (with focus on getting real results on the business, not just side/pet projects). 3) get another offer. one interview loop is just a combination of weak signals. if you get multiple offers, companies will see that as a stronger signal that you're a good hire. focus your efforts on companies you'd actually want to work at. unless you're a great actor, recruiters can tell if you're just fishing around for offers, and they're less likely to waste time getting you that max offer. 4) be a good person. make sure you're coming off as someone your interviewers would want on their team, and your recruiter would want to help out. 5) don't give up! background: 8 years progressively senior experience in data roles at various companies. moved into tech 4 years ago from big name non tech company edu: undergrad degree in social science field, no grad school but lots of regular reading on the ds industry and statistics

Thank you so much 🙏

Op here (had to create new acct). This is what I was worried about, but the only negative posts on Blind I see are primarily from juniors. Definitely agree with you on pay signaling second class citizen. Fb tries to sell the role as high impact and that the org is very "data driven" but I was pretty unimpressed with the work they told me about as examples. The team doesn't seem to really understand the causal inference work they're doing, and even the metric definition work seemed a bit shoddy. Wonder if it's survivorship bias, and the stronger DS end up working elsewhere for higher pay. 325 is by no means a terrible offer, but companies with less money than FB are paying 500+

There are many more things that I have seen at Fb which are red flags On your exp with scientists there, probably that's a bad sample .... I hv seen competent ppl there too as some ppl join for brand value

@twitter can't tell if you're joking, since your account says you're already working at twitter. if not, feel free to pm me! i think my only real concern about twitter is company outlook and fb being more stable, but twitter ds > fb ds

Yes, working at Twitter currently and have an offer from FB. FB is giving me just 10% hike. I don't have much info about DS role but feel free to dm me. My dm have exhausted.