Setting your data science team up for success: 3 critical considerations
Presented by Anaconda
In 2012, “data scientist” was famously deemed the “sexiest job of the 21st century,” with anticipation that the demand for talent would quickly outpace supply. Organizations raced to add “data-driven” to their mission statements, and data scientists found themselves at the center of talent bidding wars, commanding formidable salaries that further fanned the flames of the hype.
Alternatively, some companies tried to jump on the big data bandwagon by rebranding their business analysts or data managers as “data scientists,” giving a new name to professionals tasked with maintaining the same dashboards and pulling the same metrics as before.
Since then, data scientists have become far more common in the business world, but many organizations still fall victim to the misconception that data science is a silver bullet for any and all business problems. Businesses that hire data scientists often neglect to establish the best practices needed to position them for success. In many cases, these organizations will try to force their data scientists into a single function –business analyst, data manager, software engineer, etc. — failing to take advantage of the hybridization that makes data science unique and valuable.
Data scientists are hybrids: neither fully “business” nor fully “technical,” they combine elements of both, along with the principles of classic scientific inquiry, to offer unique value to the organizations they serve. This is one reason why we see such variety in reporting structures for data scientists, who may find themselves sitting in the IT org, operating on the business side, or working in dedicated data science centers of excellence. Any of these organizational structures can succeed, but only if leadership is able to integrate all facets of the data scientist’s role. Only by empowering data scientists to embrace their hybridized capabilities can businesses reap the full benefits of those skills. Here are three critical considerations to do just that.
1. Data scientists seek impactful work
Data science enables organizations to act more strategically by leveraging their data. To do this, data scientists must be empowered to create lasting business impact. However, according to our 2020 State of Data Science report, 41% of data scientist respondents reported that their teams could only sometimes or rarely demonstrate the impact data science has on their company’s business outcomes.
One way organizations can help their data scientists provide real business impact is by equipping them with the necessary domain expertise. Data scientists should be onboarded into the institutional knowledge of what the business does and the context in which it operates, so that they can apply their skills more effectively.
2. Data scientists want to explore
The scientific process is about challenging accepted knowledge and testing new hypotheses, and data science is no exception. Data science often centers on discovering the “unknown” unknowns, which can unlock tremendous value for how an organization can explore product or business decisions. This is a key differentiator between business analysts and data scientists: the former answer known business questions with data, while the latter examine data to find new patterns and questions to be asked.
To make these impactful discoveries, data scientists need space to explore. In fact, data exploration is a critical early step in the data science lifecycle, allowing data scientists to get up close and personal with the data they’ll be using. This process provides them with their first insights into the patterns and biases embedded in that data and enables them to form their first hypotheses while thinking through the queries, models, and features they’ll want to implement. When data scientists first approach a new problem or question, they may not know exactly where their explorations will take them, and that’s okay; in fact, it’s one of the advantages of their skillset.
3. Data scientists need innovative tools
At the forefront of a burgeoning space, data scientists need access to a diverse selection of cutting-edge tools that can facilitate their explorations, rather than restrict them. Unfortunately, too many organizations demand miracles from their data scientists while equipping them with little more than a Tableau login and a copy of Microsoft Excel. Today’s machine learning workloads need both innovative software and powerful hardware. For the former, open-source tools have become the foundational building blocks for innovation in data science, embraced even by traditional enterprises looking to equip their data scientists with the latest and greatest tools.
We’ve seen a variety of partnerships emerge between open-source providers and hardware makers to ensure that data scientists are not limited by compute power. One example of this is our own partnership with Intel to enable data science teams to operate within the bounds of IT without sacrificing enterprise governance or resource conservation.
Towards the next phase of data science
As data science continues to mature, it’s time for the business world to align on what an established, mature data science practice can be. Data scientists have an opportunity to distinguish themselves as a unique role that drives strategic transformation wherever it is applied. To seize this opportunity, organizations must embrace the hybridization of the role, providing their data scientists with the opportunities to make real business impact, explore unknowns, and use the most innovative tools available. It’s only then that data scientists will be able to usher in a new era of data-driven thinking.
Sponsored articles are content produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. Content produced by our editorial team is never influenced by advertisers or sponsors in any way. For more information, contact [email protected]
Source: Read Full Article