Mailbag: How to Define a Data Team's Vision and Roadmap

[ leadership datascience career 📬 ] · 3 min read

E writes:

Hi Eugene,

I really liked this post btw. I am heading into a team lead role and I would love to buck the trend in setting a proper roadmap and vision. Now the problem is that my manager doesn’t have this and I have done it for my own area but it’s a little challenging thinking about a small team.

Do you have any resources/books/articles/people that could help me in this journey in being a data leader? I would really appreciate your help here.

Thanks,
E

Hey E,

Congrats on the increase in scope! And I think it’s great that you’re taking the initiative to plan for your team’s roadmap and instead of accepting the status quo.

I can’t think of any specific books that relate directly to being a data science leader or setting roadmaps, but here are a few books that might be useful in your situation.

  • The First 90 Days: Useful guidelines on how to changing roles or getting promoted into a new role.
  • An Elegant Puzzle: Discussions on engineering management, which I think is relevant to data too.

Specific to setting roadmaps, here’s how you could try doing it:

  • Schedule some time with key stakeholders of your new team. (These are the people that your team would serve.)
  • Ask them: What keeps you up at night? What are your key KPIs? What can my team do to help you meet those KPIs or improve your productivity?
  • Then, evaluate those key problems via cost-benefit analysis. Also, are there multiple problems that are downstream symptoms of a root cause, and is fixing this root problem an 80/20 effort? Finally, are any solutions also applicable and usable across the org and act as a multiplier? More on evaluating problems here.
  • Identify the top 3 problems (or as many as your team can handle), and write one-pagers for them via the Why, What, How approach. Share them with stakeholders to get their feedback and buy-in, and then you can add it to your roadmap.

Note: You don’t have to be the only person working on the above! Delegate some to your team and mentor them to help them grow. Also, If the above seems like a lot, yes it is. Nonetheless, I think it’s an effective use of time. Measure twice, cut once. Also see Amazon’s Working Backwards. That said, tweak the process to meet the timelines of your org, be it a startup or mature company.

All the best,
Eugene


Have a question for me? Happy to answer concise questions via email on topics I know about. More details in How I Can Help.


Share on:

Browse related tags: [ leadership datascience career 📬 ]

Join 9,100+ readers getting updates on machine learning, RecSys, LLMs, and engineering.