With all of the buzz about data science and companies embracing the value that data science can bring to a business, organizations are starting to ask the next logical question: What resources are needed to fully reap the benefits in this area?
Some are evaluating whether to spin up their own data-science teams, bolt-on open-source and SaaS platforms, or just wait until the dust settles so the landscape is clearer. From where we sit, a good strategy involves a mix of custom and off-the-shelf approaches that will produce organizational alignment and business imperatives for rapid innovation to outpace competition and incremental revenue growth.
For companies considering how to best scale data science, here are a trio of key points to consider:
1. Data science is about knowledge, not about the degree: No specific degree or work experience qualifies someone to call himself a data scientist. One emerging view of the profession is that a data scientist possesses three specific skill sets: subject-matter expertise, applied statistics and math, and hacking or computer science experience. Together, these skills enable the data scientist to ask the right questions, design experiments and apply statistical methods to business problems, and develop and validate prototypes.
No single candidate can have world-class skills in all three areas. Organizations with the most effective data-science programs assemble teams from diverse specialties and provide strategic access to the internal and external resources as needed. The team is then aligned with executive sponsorship and focused to meet the business imperative and metrics.
2. Organization alignment matters: The most effective innovators instinctively know that innovation isn’t a job or a role–it’s an organizational strategy that requires the right amount of sponsorship and independence in order to drive results. Data science is similar: In order to organically drive innovation and revenue, this team must be definedin terms of objectives, yet empowered to pursue its own independent lines of inquiry. It’s important to have the data-science experts define the actual tactics for execution because they are the ones with the most industry insight and knowledge.
3. Don’t reinvent; revolutionize: Any data-science team has to learn before it can lead. One of the biggest questions in focusing this team is fundamentally a build-versus-buy decision: Which projects should a data-science team take on internally, and which should leverage outsourced or partner resources? While exceptions exist for every rule, in general, your data science team should focus on leveraging subject-matter expertise at identifying new opportunities and innovating on your company’s core value proposition. For first-use cases, outside partnering will save time, generating learnings and insight for the next instance. Where it makes sense, allow your innovation leader and data science team to focus on revolutionizing rather than on reinventing.
Ready or not, data science has escaped the lab. The strategic importance of being able to compete and win on analytics is becoming increasingly clear. Businesses of all stripes are working to leverage their vast amounts of customer data to generate meaningful–and monetizable–insight.
In short, there’s no time like the present to get started with data science.