In my experience, some of the most talented analytics professionals I’ve managed were ones that had intimate knowledge of the system limitations required to meet customer needs. These individuals came from a variety of roles, some from engineering, and others from customer service roles. Their strength was in forming specific hypotheses to pinpoint customer experience issues and then leveraging their curiosity to do whatever it took, including learning new statistical techniques and programming languages, to address the project need. When these “business analyst” types were mixed with more traditional data science talent the result was almost always 1+1=10.
What we like about Bernard Marr’s article last week about the Citizen Data Scientist is that the concept starts to solve for the so-called “shortage of data science” talent by empowering qualified people in the business to answer core questions. During their 2012 IPO road show, Tableau CEO Christian Chabot said, “We believe there’s a tremendous opportunity to help people answer questions, solve problems and generate meaning from data in a way that has never before been possible. And we believe there’s an opportunity to put that power in the hands of a much broader population of people.” Tools like Tableau are certainly where they need to be right now to enable this sort of evolution Marr describes.
There are many positives about the Citizen Data Scientist concept, starting with the assumption that the significant upfront investment to transform core business data into a tangible asset is complete. Additional positives include:
- The empowerment of pockets of the business with large domain expertise to hypothesize about the business and develop more actionable insights with direct access to data.
- The freeing up quantitative talent to answer more complex business questions.
The end result should be a flourishing of insights and a more efficient use of resources. Business partners will have greater control regarding how quickly they answer their key questions while data scientists will spend more time flexing the quantitative skills they’ve built over the years by solving more complex problems.
The primary obstacle to starting down this path is the roles/individuals intent on controlling the data and limiting data access. Companies with centralized analytics functions can be entangled in this type of command-control arrangement which forces business partners to queue up “analytics questions” that are better answered by members of their own team. When faced with this dilemma, centralized analytic teams have a choice. They can be stewards for their internal partners and answer sets of questions that tend to deliver incremental innovation or they can focus on larger strategic initiatives that don’t always align with partner needs and can cause unnecessary friction.
Over the next several years as technical resources have greater ability to serve data in a variety of ways, greater innovation will come from companies that look to share raw data access broadly to allow for “local answers” to sprout up to specific business problems. Democratizing the Data doesn’t just come from tools like Tableau or easier data query front-ends. A different data sharing philosophy is required to allow all business functions to participate in analytics.
Try deep learning using MATLAB