Definitely depends on the company, but centralized teams can often become a bottleneck that creates a kind of "shadow IT" pattern but for analytics work, specifically. Where the teams of interest - CS, Product, Marketing, whatever - go rogue and take things into their own hands rather than using analysts in a centralized team. I've seen success with all different team structures, but the common theme is usually keep the data team + infrastructure consistent with the way the rest of the org is structured
Ones that resonate most (read: I learned the hard way): 1, 11, 14, 16
I haven't had the same experience with 7 - care to share more?
Definitely depends on the company, but centralized teams can often become a bottleneck that creates a kind of "shadow IT" pattern but for analytics work, specifically. Where the teams of interest - CS, Product, Marketing, whatever - go rogue and take things into their own hands rather than using analysts in a centralized team. I've seen success with all different team structures, but the common theme is usually keep the data team + infrastructure consistent with the way the rest of the org is structured