The data science field is growing — and it’s not expected to slow down anytime soon.
According to data from the Bureau of Labor Statistics, there were 63,200 people employed in data scientists and mathematical science occupations in 2020. Between 2020 and 2030, there are expected to be nearly another 20,000 new jobs added to the field — a nearly one-third increase in the profession.
“Data is the new oil,” Clive Humby, a noted data scientist, said in 2011.
If that’s the case — which, with the growth of the profession and its near-ubiquity in tech companies today, it seems to be — the world needs more people to make sense of the constant stream of data.
To respond to the growth, data teams at companies across the country have had to scale their teams with perhaps just a few members to ones with tens and dozens of members. And the numbers don’t lie: As of this writing, there are more than 460 roles on Built In for data scientists across the country.
In conversations with two local data leaders, Built In Chicago discovered that data scientists need to not only have the technical wherewithal but excellent soft skills to collaborate with larger teams. Onboarding and initial training are important, but so is continual learning and keeping up to date with new tech and workflows.
What they do: Peapod Digital Labs is the e-commerce engine of Ahold Delhaize USA, one of the nation’s largest grocery retail groups.
When it comes to scaling your data team, what are the most important hiring considerations and why?
Analytic and coding skills are table stakes. Aside from those, I look for people who enjoy being part of a team and are good communicators. We are a large team and often work on the same problem from different angles, so collaboration is key. I value domain expertise in retail, particularly grocery, and preferably those with an online and physical store presence.
Oh, and I love people with strong opinions on supermarkets or grocery products! We analyze customer behavior to create a better experience and it helps to remember your frustration that time your favorite product was not on the shelf.
Ahold Delhaize has more than 2,000 stores in the U.S. and a thriving online grocery business — that’s an immense amount of data.”
On the technical side, what steps have you taken to make sure your tools, systems, processes and workflows are set up to scale successfully alongside your team?
Our data landscape is continually evolving, whether that be the ability to pay online using EBT, new tools or a new process to deploy models. It’s key to encourage curiosity and innovation in the concept stage, but successful scaling requires rigor. For example, we might start developing a new model in R, then switch to PySpark for the final model. Code must match internal standards, be unit tested and validated, and tuned for performance. There are clear criteria for handover to the production team. Ahold Delhaize has more than 2,000 stores in the U.S. and a thriving online grocery business — that’s an immense amount of data every day before we start overlaying promotional offers and marketing communications.
What’s the most important lesson you've learned as you’ve scaled your data team, and how do you continue to apply that lesson?
To appreciate the value that each person brings to the team and to continue to invest in their development. And yes, that covers learning new skills, attending conferences and regular benchmarking against the market. But it also means making sure every voice is heard.
Collaboration and sense of belonging became more important and challenging as we rapidly grew the team. Actually, remote working helped us. We set up a chat channel for the team where everyone shares their accomplishments, data questions, thoughts, and interests — and we’ll keep using it even when we are finally back in the office. I’m not sure we even notice that we are spread across five states!
What they do: Beyond Finance is a financial services company that helps users eliminate debt using a range of software products.
When it comes to scaling your data team, what are the most important hiring considerations and why?
Under pressure to scale up quickly, there’s a temptation to relax standards in order to get candidates through the pipeline faster. But that’s a short-term tradeoff that can result in long-term headaches due to hires that turn out to be bad fits for the team.
We’re committed to remaining rigorous in our recruitment process. For our analytics team, data analysis is coupled to strategy. How can our analyses produce meaningful insights that help attain our company goals? When recruiting, we look for individuals that have a track record of using data to drive impactful results. Top-notch data and SQL skills are a must, but individuals who thrive on our team are also strategic thinkers.
When recruiting, we look for individuals that have a track record of using data to drive impactful results.”
On the technical side, what steps have you taken to make sure your tools, systems, processes and workflows are set up to scale successfully alongside your team?
We’re always asking if our tools and processes are still working for us, and if not, upgrading them to something better. A 15-minute daily standup may work great for a team of six. But when the team doubles and the meeting stretches to 45 minutes, it’s time to revisit. What is the purpose of this meeting? Is this still the best way to accomplish this? We’re not afraid to scrap things that are no longer working as originally intended.
Small teams can get away with less structure, and data analysts might be fine managing their own SQL code. However, a code management solution like GitHub is an essential tool as teams scale to create a single source of verified code.
Responsibility for evaluating and upgrading tools and workflows falls on our whole team. The goal is to ensure tools are always in service of the team, and not the other way around.
What’s the most important lesson you’ve learned as you’ve scaled your data team, and how do you continue to apply that lesson?
Initial onboarding is crucial for getting new team members up to speed and able to contribute quickly. It’s very easy to overlook things that experienced team members take for granted. A new team member could spend several hours struggling to get connected to our data warehouse when the answer is as simple as needing to connect to the VPN.
We’ve established a list of important resources shared on day one to get need-to-know information disseminated quickly. Additionally, our hiring managers utilize an onboarding checklist to ensure access to essential tools is set up before new team members start so they can hit the ground running.
