Ready to learn Data Science? Browse courses like Data Science Training and Certification developed by industry thought leaders and Experfy in Harvard Innovation Lab.
Losing sight of the BIG picture – As data scientists, our job is to “extract a signal from noise,” and be able to glean insights hidden in the data which will ultimately have an impact on business. Hence, the inability to zoom in and out of the problem being solved and having an obsession with accuracy and depth at the cost of breadth can cause projects to fail. As an example, you may get 85% accuracy from your models in a month but moving from 85% to 90% could take you a year. For every iteration- is it really worth the cost and effort to work towards being super detailed and accurate? To find the right balance of depth and breadth, get feedback from your stakeholders quickly. Engage them early to ensure your recommendations generate value for them and fit into the larger scheme of things.
Lack of engagement with key stakeholders – The right people, processes and culture is bedrock for building viable frameworks and infrastructure for data science. If you hear quotes like – “we have some data science folks but no one really knows what they do,” then it signals a lack of executive sponsorship and engagement with stakeholders. Just hiring data scientists is not enough – this talent needs to be integrated into the existing organization and new structures that enable value creation, need to be invented if required. One approach could be that data science professionals are tied to business units, and share responsibility for BU performance. Whatever approach is taken, and like most strategic initiatives – success is driven by executive sponsorship and buy-in from senior management.
Putting the ‘How’ before the ‘Why’ – When you begin with scoping a problem, freeze the “what” and “why” first. Problem formulation meetings that start with ”how” become inevitably shortsighted. This can happen a lot if you have a group of bright, technically inclined folks raring to dive into the latest tools and technologies at the word go. However, good problem statements always capture the “trigger” which caused the problem to surface. Putting the data before the question can cause you to be lost in the enormity of the data and tools being used. This can also lead you away from solving the right problem.
Not solving the right problem – One of my favorite quotes from the statistician John Tukey captures the essence of this issue perfectly – “An approximate answer to the right problem is worth a good deal more than an exact answer to an approximate problem“. This challenge manifests itself in many forms – if you don’t know what problem to solve, how does it fit in the overall scheme of things? Avoid this by asking for a use case and getting diverse opinions. You need to identify the problem and solve it in a manner that will result in the right outcome.
Hiring Data Scientists who are Unicorns – Data science is a field that requires an interdisciplinary skillset. You need to be good at math and statistics, which yields a foundation of methods to analyze and interpret data. Domain knowledge is required to understand the data and the (business) processes that shall benefit from the analysis. Coding is a prerequisite to bring the theory to action. It is hard to find one person who has all of these capabilities. Unicorns don’t exist, or are extremely rare to come by. Hence, instead of struggling to hire them, build diverse teams. Data science is a team sport, and constructing a team with a strong combination of these skills will help you avoid getting into a hiring quagmire!
Originally posted at insideBIGDATA
