Need to present data in such a way that anyone can understand it: Though a data scientist should be well-equipped with all the technical and machine languages like R, Python, etc., but he should present the data in a facile and simpler way so that even a layman can understand the insight from the data.
A data scientist should never show a regression analysis or a plot from R because only a few people have an adequate knowledge regarding these. Rather he should present the data in a story telling way which consists of simple slides and visuals instead of numbers. Visualizing and communicating data are equally important, specially for the nascent companies who are making the data-driven decisions for the first time or the companies where these professionals are viewed as people who help others making data-oriented decisions.