The most important people in any organization are those who have a solid mix of technical skills and business acumen.
The most important people in any organization are those who have a solid mix of technical skills and business acumen. It does not matter where they sit but they need to be able to collaborate with others. Considering that most data scientists come from technical backgrounds and academia, how do they acquire the "business acumen" before being recruited ? Maybe companies should accept that this is something that a data scientist can learn on the job and as a matter of fact, no entity is better equipped to teach about business than the companies themselves that require the services. If there is one skill to put emphasis on recruiting a data scientist, it's the ability to learn new concept and techniques and put them in practice reasonably fast. Not everybody that works in data works in the business field. For data analysis is vital to understand the domain knowledge in order to understand the history your data is telling. As with any engineering or science related job you need to learn to think and be intelligent in order to be good at your job. A data scientist should be able to take any problem and translate it to the abstract domain of analytics and data science.
Data science is multi-disciplinary. The skills required include math, statistics, programming, data management, communication, visualization, domain knowledge and soft skills. Business knowledge is as important as data because if you are not able to interpret it in a business understandable way then there is no point projecting or predicting something for business. Data scientist must know the technical part, but business knowledge helps in communicating data to management and business owners which are very important as if most of the decisions are taken in account data then it must be communicated properly. Data knowledge is a waste until you know where and how to express and implement it, and business domain knowledge is equally as important as technical knowledge. Domain knowledge will help the data scientist know when the data doesn't make sense. There are times when the data looks right based on technical stuff, but events in the field are very much different. Very few individuals who even come close to embodying the range of skills and experience required. Effective data science results from close collaboration between technology and domain experts. The goal of a data scientist is to understand the business and the data and make a bridge between them, both by programming solutions which take advantage of the data. But equally important by making other people understand this understanding. Real data scientists do not sit in an ivory tower debating the best algorithm but understand the needs of normal people and find together with them a viable solution.
The language of visual communication. Communication is paramount. A data scientist should focus on core math/decision science/computation skills and collaborate with the business analyst for domain knowledge. Collaboration and communication are important. No matter how they are arrived at, the insights gained by data scientists must be implemented as actions and that takes convincing a group of non-data scientists that it is important and will show results. Communicating that is a skill in itself. It is not limited to data scientists, but all talent employees who are employed by businesses. If you are closely aligned to the business, you should be able to contribute more intelligently to the business problem you are being asked to solve. The key challenge is that management must integrate developers/data scientists to their business teams. Easier said than done, but this is the fundamental issue. Business people don't "get" tech, and Techies don't "get" business.... There are exceptions, of course, but the skills can be cross trained to a degree, the key is encouragement from both camps to get people to enjoy and benefit personally from the crossover. Companies that have issues with this are classic examples of IT/Business finger pointing and they will be fossils only remembered in history books if they do not embrace this necessity in the modern world to encourage cross-pollination.
Build a data-driven and high performing team to improve analytics maturity. Data Scientist will definitely contribute to the business growth from data mining, data analysis, and modeling. The data science community need to realize that they are there to support the decision makers in the organization - data scientists are not decision makers and don't set business strategy, they supply the ammunition for others. Unless the problems you are trying to solve are purely technical, you will want heterogeneous teams that combine business, technical skills, and theoretical knowledge. These interactions will then ensure the data scientist that come with a more technical or scientific background will learn about the business side of the problem. A company should be a team, that some people in that team must understand business, and some people are masters of other disciplines. How that team pans out, how big it is, will be a function of the size of the company.
Clearly if you can find someone with a business sense and is good at understanding data, then you are very lucky - but in general, those disciplines don't go together seamlessly; so the key is communication and how that team is able to use and put into practice and provide ROI to the company. So the team must be good at business, and the leaders of that team must be good at directing the communication to get the best. And there is always room for communication improvement. If organizations don't communicate well, they fail.