How to Survive as the Lone Data Scientist in an Organization

It’s the one thing you can do that will make your life easier

Harpreet Sahota
Artificialis

--

Photo by Ray Hennessy on Unsplash

I was speaking with a mentee of mine who recently joined an insurance company as their first data scientist.

And one of the cold realities they were faced with was the realization that the company didn’t have proper data management or governance, no data architecture, and no skilled data warehousing experts.

It lacked data maturity in every sense of the term.

Which makes our jobs as data scientists…tough (understatement of the year).

But the course of the conversation hinted at something deeper, and I realized what the question was truly about.

It wasn’t really about data warehousing, as much as what to do when you’re a lone data scientist.

This is a position I’ve been in a couple of times in my career, and one that more data scientists than you think find themselves in.

And that is: The Hero Data Scientist.

And when you’re that Hero Data Scientist, what do you do?

The first thing you need to do is recruit people.

So what does this mean?

This means you’ve got to find people that have aligned interests with you.

If you see that the company doesn’t have its data maturity to a level that you need it to be at to be effective, then you’re in trouble as a data scientist.

Because as an individual contributor in the company, you’re not going to have enough leverage to get the company to change the way they deal with data.

Now, you could try to fight this battle by yourself and maybe, eventually, you might win.

If you’re loud enough, if you’re persuasive enough, if you have enough leverage…maybe you can win this on your own.

But, you don’t really want to do that.

No one wants to go into battle alone.

So, you recruit people to your team and find people that also would benefit from the improvements that you want.

If you want to improve data warehousing, maybe you think the company should hire someone to just focus on data architecture, governance, ETL, whatever it is.

You’ve got to think who else would benefit from that happening if that happens? Who has aligned interests? Who also would like that to happen?

Photo by Surface on Unsplash

Maybe the first people you speak to are line level employees, who are actually out there doing the work.

Talk to these people and sell them on how awesome it would be if you had someone to help organize this data and to automate the ETL process.

Sell them on how awesome it would be to have process that is clean, reliable, well-designed as opposed to something that is just thrown together.

Photo by Nathan Dumlao on Unsplash

Create a shared vision of a future that you could have if these outdated, old legacy solutions weren’t holding you back.

Next, you want to find an authority that also has aligned interests.

Tell them you’ve spoken to people on their team.

Tell them you came to a shared realization that if you had a person that was going focus entirely yon data warehouses, people on your team would free up 30 percent of their time and be so much more effective.

Paint a picture for how it would help their department, how it would decrease their costs and increase their revenue, how it would free up time to work on that other project that they’re much more passionate about and really want to focus on.

Once you find that authority who is also going to benefit from this, you ask them to sponsor you or your request.

Because you need their help.

Because you think that they’re the one person that can really make this happen for the team.

And if they help they’ll be much more successful and more importantly make our customers happy which will make the company more successful.

When you’re the lone data scientist success depends on recruiting people to your side, seeing things from their perspective, and thinking about how your work or what you want done can benefit them.

And then you can appeal to authority and have them sponsor that change for you.

That’s it for this one!

Let me know what you think and if you have any advice for those lone data scientists. Feel free to comment below.

If you’re the lone data scientist in an organization, don’t worry, you’re not alone! I’ve got a free, open Slack community where we’re all happy to help and trade war stories with you!

And remember: You’ve got one life on this planet, why not try to do something big?

--

--

Harpreet Sahota
Artificialis

🤖 Generative AI Hacker | 👨🏽‍💻 AI Engineer | Hacker-in- Residence at Voxel 51