There are different ways of organizing a data science team. In some tech companies, data scientists are all on one team and each data scientist works like a consultant with various product/business units. In others, like Facebook and Jana, data scientists are embedded into the product teams – they are members of the growth or retention or fraud teams and work alongside PMs, engineers, and designers.
I’ve worked in both types of models, and there are definitely pros and cons to each – on Google, you can find a wealth of articles on this topic. I personally have found that the embedded model works really well for data scientists in product teams that are moving very quickly.
As a data scientist on the growth team, my goals and objectives are aligned with the people I collaborate with every day. And should those objectives change, I can adapt quickly with them. Also, with time, my understanding of what drives growth or retention in our product grows, and that helps me be more productive with how I look at data and share those insights with my team.
Even if you’re embedded on one team, at Jana we still have weekly huddles and a Slack channel with other data folks in the company to share what we’ve learned, different ways to tackle problems, and learn from one another.
How do data scientist/engineer/analysts work with your product teams?