Pros
It looks good on a resume! You are influencing products that have an existing user-base of millions. They encourage movement within the company, which means that once you get in you can experience lots of different patterns and solve lots of different problems.
Cons
Management practices are really inconsistent, definition of what it means to be a data scientist varies a lot by team, and tooling/technologies are not always fit for purpose. The first team I was on didn't seem to use any kind of database; the second team I was on had some of the worst spaghetti code I've ever seen. The tasks I had were not typically data science related, and I suspect that given my background and my gender I was getting relegated to less technical tasks, even though I've excelled in technical roles in previous companies.