You know that you need to harness your data to stay competitive, but where will you find these data scientists? Before rushing to hire an army of computer science candidates with doctorates, understand what successful data science teams do.
Data scientists write code that applies complex algorithms to analyze and transform data into actionable insights, and help IT departments determine optimal structure for data storage. But transformation and storage are only pieces of the larger data life cycle.
Data science teams also need to do the following:
- Identify with business stakeholders what initiatives will solve a real need.
- Visualize and communicate the data intelligibly to business stakeholders.
- Design intuitive data products used by business stakeholders.
- Advise business stakeholders on the political and legal context for the data.
Seek diversity
Finding an individual with expertise in everything is nearly impossible, but you can combine candidates that have the full breadth of knowledge with expertise in one or two of them. Diverse teams will yield the greatest return for your organization.
Recently, Booz Allen Hamilton hired a diverse group of experts to help synthesize a pharmaceutical client’s adverse-drug-reaction, social media, and lab and molecular data with research notes to reprioritize their drug research pipeline. For smaller companies lacking resources to hire or develop a team, use the questions you hope to answer to guide hiring. Sometimes the most challenging phase of the project has nothing to do with the computer programing.
Fit the team to the challenge
Stylitics, a startup that uses individuals’ closets to help designers and fashion houses predict fashion trends, wanted to answer the question: How can we learn the last 50 things you bought and the last 50 things you wore?
The company recognized that the most scalable way to capture this data was to provide immediate value back to users and create a digital closet that tracks what they wear, when they last wore it and so on.
The biggest challenge was data collection, not complex programming. In your business, the solutions you seek may not all require complicated engineering solutions, but rather a team or individual with the creativity and diversity to find it.
So what should you do to help your company deal with your own data challenges? Some next steps:
- Look internally — Often, it’s easier to provide training to someone already familiar with your business’ goals and data. You may discover that your next data scientist is not in tech.
- Develop a diagnostic — HR should develop internal diagnostics to identify appropriate candidates. Additional guidance can be found in reports from O’Reilly Media and Booz Allen Hamilton.
- Keep them sharp — The field of data science changes rapidly and you should provide opportunities to learn new techniques and skills when necessary. Master’s degrees, including a degree in information and data science, and other resources are increasingly available.
- Give them company — The best solutions come from teams with diverse experts. Starting off with one data scientist is fine, but try to have a plan to grow the team and listen closely to feedback from your initial hire(s). ●
AnnaLee Saxenian is dean and professor in the School of Information and professor in the Department of City and Regional Planning at the University of California, Berkeley. Her most recent book, “The New Argonauts: Regional Advantage in the Global Economy,” explores how the “brain circulation” by immigrant engineers from Silicon Valley has transferred technology entrepreneurship to emerging regions in China, India, Taiwan and Israel.
Link up with AnnaLee Saxenian:
LinkedIn: http://linkd.in/1eysatN
