5 Ways to Find the Data Scientist of Your Dreams

To succeed in the data decade, businesses need to empower their data scientists to do the extraordinary. Yet many struggle to do so. Here are five ways to find the right data scientists and help them thrive.

By Donagh Buckley, Strategy Leader Data Management, Dell Technologies and Yeshaswi Bharadwaj, Strategy Lead and Business Analyst for Intelligent Data Management, Dell Technologies

A couple of years back, experts at an Institute for the Future workshop estimated that about 85% of jobs needed in 2030 hadn’t been invented yet. Data scientists are at the bleeding edge of this new workforce in the way that their role has evolved in recent years. It has unmatched potential to fundamentally impact everything from innovation, to the shape of business, university programs and more.

That said, as an emerging field, there aren’t enough data scientists for the volume of work which is problematic given that data is now the lifeblood of a modern organization. To compete in the data era, businesses need data scientists to help them collect, audit, tag, clean and analyze an exponential growth of data and then put that data to work, whether that’s by identifying opportunities to optimize business processes, deliver more efficiencies and personalize customer offerings.

However, even when companies can boast an enviable line-up of talented data scientists, much of their time is being squandered on low value tasks.

To address the dearth of high-skilled data scientists, and our perception that many of their skills are being misdirected, we’ve compiled five tips to help businesses recruit and nurture these roles.

1. Start with the job spec

Data science is an inter-disciplinary field related to data mining, machine learning, and analytics. As such, the job title ‘data scientists’ has become the de facto umbrella term for a myriad of emerging tech roles.

When it comes to building out their teams, businesses will need to get a lot more granular and scope out exactly which data roles they will need and for what.

A starting point would be working with stakeholders to crystalize their goals and determine how data can be used to achieve them.

In doing so they’ll need to consider each stage of the data lifecycle: from finding and gaining access to data (a huge drain on time), to designing data modelling processes, creating algorithms and predictive models, extracting the data the business needs, and ultimately helping to analyze the data and share insights with peers. Every phase needs to be properly resourced and supported.

2. Automate the mundane

In general, grunt work can be automated, whereas higher value work requires human discernment. This applies to data work too.

Too often data scientists are overloaded with the exhausting and laborious task of trying to uncover what data sources are available to them, how to gain access to that data that can often be spread across multiple siloes, and then waiting for others to manually action email requests or tickets, etc.

According to recent third-party research, data scientists are often spending their time on low-value, ad hoc reporting requests that grow and “spiral into never-ending projects.” As such, many feel their work is under-appreciated by company leadership.

Frankly this is crazy. Data Scientists have a pivotal role to play in breaking data out of silos and driving profound change to a company’s bottom line. Sure, they can drive technical implementations, but they also need to be liberated to envision what can be achieved with data and how to make this happen. With so many complex rules around data governance, data sovereignty and data privacy, they also need to put the necessary safeguards in place to build trust in company data which plays a key role in building a data-driven culture.

Their time and skills are invaluable assets and should be protected as such.

3. Scout out talent

Given the shortage of talent, businesses could look at their current employee base. Hosting hackathons with wide cross-section of your workforce might draw them out. Creating 360-degree review processes that can identify team members proficient in areas like computer science and math would equally be a good start. Additionally, you could offer re-skilling programs for developers or other numerate disciplines to become data scientists.

Try not to get too hung up on qualifications. Given the pace of change, many of these qualifications could be obsolete in a few short years anyway. Rather, you’re looking for an aptitude to learn in the moment, fail-fast and then succeed.

Of course, you’ll need to sell the role to them, but they probably won’t need too much convincing. Chances are, they would want to add more skills to their bow – particularly when they see that these roles are highly valued, well rewarded – and here to stay.

4. Democratize data access

Linked to the above point, to expose your wider workforce to the joys of data science, you’ll need to broaden access to the data, by encouraging collaboration across functional teams of technical and non-technical users and enabling them to self-service.

Naturally, the data needs to be relatively easy to find first, of a high quality and production-ready – to the untrained eye at least. But this can be achieved with machine learning automation technologies, and by standardizing key processes, such as how to transfer learning from large data sets into smaller ones. In essence, data science shouldn’t be elitist and even if you don’t set out to enlist new recruits internally, democratizing data access will enable you to create a more data-friendly support group for your data scientists.

Once cross-functional projects bear fruits, don’t forget to promote and celebrate these successes across the organization to encourage future participation from more employees.

5. Provide the right tools

There’s a famous saying, “a man is only as great as his tools”. That’s still true – though the same applies to men and women alike. To create a productive work situation in which data scientists thrive and deliver game-changing insights, they need to be properly equipped with intuitive data management tools that can aid the holistic data management journey.

People and machines together, is a highly effective duo. But each part requires optimum conditions to function well. We can handle the technology, but businesses will need to look at the people and culture part. It doesn’t end with recruitment. To succeed in the data decade, businesses need to empower their data scientists to do the extraordinary and unveil insights that will take the business in new, more profitable directions.

This article is part of the “Putting Data to Work” series. See further articles that may interest you below.